[141] COMPUTER ACCESS SELECTOR AND VOCASELECT
Barry Seeger, PhD; Rob Garrett, BTech; Hugh Stewart,
BAppSc (OT); Duane A. Stapleton, BEng
Regency Park Rehabilitation Engineering, Regency Park,
5942, South Australia; email:
barry_r._seeger@ieaust.org.au
Sponsor: Australian Department of Human Services and Health, Canberra, ACT 2601, Australia
PURPOSE--Computer Access Selector and VOCAselect are computer programs that aim to assist in the selection of alternative computer access and voice output communication devices for individuals with disabilities, which selection involves a compromise of many different parameters and generally requires professional assessment of the patient's the needs and abilities. Even when the capabilities of the patient are understood, the choice of specific products is not necessarily straightforward due to the number of them available. Computer Access Selector and VOCAselect narrow down this large number of products by feature selections made by the prospective user.
METHODOLOGY--The programs give the user an opportunity to specify the requirements of the devices they are looking for by selecting from a series of parameters. They then use expert knowledge to suggest potentially suitable products. The suggestions are dynamically adjusted as user selections are changed, providing a means for the user to explore different possibilities and trade-offs.
With one click of the mouse the user is able to view details about any given product, and see a picture of it onscreen. Manufacturer and Australian supplier contact information is also provided. The software can generate a report of the user's findings for printing or inclusion in a word processing document.
PROGRESS--Computer Access Selector version 1.0, a Windows program, was released in October 1995 at the Second Australian Conference on Technology for People with Disabilities. It was subsequently offered free for evaluation purposes and has now been distributed to approximately 300 interested individuals, professionals, and organizations.
VOCAselect version 1.0, a Macintosh program, was released in November 1995, and has now also been distributed widely. Both programs underwent alpha and beta testing prior to release and are available for $35 + $10 postage and handling each.
RESULTS--Feedback and responses to Computer Access Selector and VOCAselect have been sought from users primarily via a survey questionnaire sent with the software. The survey results are now being collated.
FUTURE PLANS--Refinements based on user comments and feedback will be made to ensure that the programs continue to be relevant and valuable tools for all who need to choose alternative access and voice output technology. A significant percentage of VOCAselect users have requested that the software be made available for Windows, and such a version is now being written for release by December 1996.
The product information contained in both software packages is expected to be updated annually. Version 2.0 of both packages with updated product information will be available by early 1997.
[142] REMOTE REHABILITATION SERVICES NETWORK
Geb Verburg, MA; Andre Wisaksana, MSc; Anastasia
Cheetham, MSc
Research in Cognitive Development Unit, Rehabilitation
Engineering Department, Bloorview MacMillan Centre
(MacMillan site), Toronto; the Telecommunication Team,
Ontario Rehabilitation Technology Consortium; email:
verburg@ecf.utoronto.ca;
ortcgv@oise.on.ca
Sponsor: Bank of Montreal through The Hugh MacMillan Children's Foundation
PURPOSE--Telecommunication technology has the potential to provide remote access to resources and expertise along with access to social, cultural and health services. Video and interactive video allows people to consult and attend meetings, telecommute, demonstrate, and trouble-shoot devices remotely, and could provide ways to train professionals or parents and to distribute software upgrades and training materials.
This project will evaluate the need, feasibility, and costs of remote rehabilitation services. We will determine the needs, opportunities, costs, and potential savings that may be realized by using multimedia telecommunication technology in Children's Treatment Centres and possibly other community healthcare environments.
Some of the arguments that reinforce the need for a multimedia remote service network are:
The Easter Seal Society (ESS) of Ontario assists with the travel expenses of parents who require rehabilitation and health care services for their children, advancing the travel funds and claiming approximately two-thirds back from the Northern Travel Grant program of the Ministry of Health (MoH). In 1993 alone, this amounted to $328,887, and includes only travel of families. It does not include travel of occupational, physical, or speech therapists, physicians, nurses, and technologists, costs which are funded by MoH. Nor does it include the hidden costs of parents, losing several days to a week of work each, since in many instances the severity of involvement of the child compels both parents to accompany their child. Finally, the hidden cost of the loss of education days to the child is not expressed in the above figure.
The lack of availability of appropriate expertise and/or the limited access to technological expertise is more likely to result in the abandonment of the device or the disillusionment of parents or teachers with this, often complex, assistive technology. Broad band information technology that is available today has the potential to eliminate much of the travel and device abandonment by allowing experts to consult much more directly and appropriately.
PROGRESS--One VISIT station has been placed with a family with 12 children with disabilities. A draft of a paper reviewing applications of several multimedia technologies in health and education is in preparation. Additional funding was obtained from the Bank of Montreal to set up a network of Proshare multimedia televideo stations in Northern Ontario. This project started with surveys of inhouse and remote clinical staff to determine immediate uses of multimedia telecommunication technologies.
FUTURE PLANS--Our team plans to apply to the upcoming CANARIE INC. competition to develop new assessment products and software-based tools that can be used in televideo rehabilitation services.
[143] ASSISTIVE CONTROL IN USING COMPUTER DEVICES FOR THOSE WITH PATHOLOGICAL TREMOR
Daniel S. Hsu, MS; Cameron N. Riviere, PhD; Nitish V.
Thakor, MD, PhD
Human Machine Interface Laboratory, Department of
Biomedical Engineering, Johns Hopkins University, Baltimore
MD, 21205; email: cyclops@jhunix.hcf.jhu.edu
Sponsor: National Institute on Disability and Rehabilitation Research, Washington, DC 22202
PURPOSE--The purpose of this research is to help those with pathological tremor use computer devices like a mouse. Tremor patients are hindered from performing everyday tasks that many of us take for granted. Handwriting is among the most difficult of activities for many persons with pathological tremor. Using computer devices such as a mouse/pen is also a difficult task. With these two difficulties in mind, the goal then is to develop a computer assistive interface where handwriting is aided as well as is general control over a computer mouse/pen.
METHODOLOGY--The assistive computer interface was devised by implementing a filtering algorithm in mouse driver software to cancel tremor movements. Tremor is modeled as a sinusoidal signal with a changing amplitude and frequency (usually higher than voluntary motion). As input is received from a computer mouse/pen, the software models the tremor at every time step and acts to cancel it. While the movements of the user may be unsteady, but when processed by the computer they become smoother and more controllable on the computer screen as cursor or arrow.
PROGRESS--The filtering algorithm has been implemented in computer mouse drivers. Numerous tests on pathological tremor subjects have been completed with favorable results.
RESULTS--The filtering algorithm has demonstrated its effectiveness. In qualitative tests, the handwriting of a tremor subject has increased in legibility when the algorithm has been implemented on or off-line. The filtering algorithm has also improved performance in various on-line target tracking tasks that measure the algorithm's effectiveness quantitatively.
FUTURE PLANS--Further quantitative results are desired for a complete and comprehensive set of data. Also direction of the research is being geared toward handwritten optical character recognition (OCR). OCR is the ability for a computer to recognize and read text or handwritten letters. OCR can be used as a quantitative testing tool as well as a general assistive interface.
[144] CONSUMER INNOVATION LABORATORY OF THE ROBOTICS RERC
James W. Fee, Jr., MS; Richard Mahoney, PhD; Whitney
Sample, Randy Glass
RERC on Rehabilitation Robotics, Applied Science and
Engineering Laboratories, University of Delaware/A.I.
duPont Institute, Wilmington, DE 19899; email:
fee@asel.udel.edu
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 20202; Nemours Foundation, A.I. duPont Institute, Wilmington, DE 19899
PURPOSE--The Consumer Innovation Laboratory (ConLab) is a consumer-led research effort intended to draw upon the experiences of consumers with disabilities to establish designs for robotic devices that will enhance their independence.
METHODOLOGY--This project is an exciting and novel demonstration of consumer empowerment. It exposes consumers to the engineering of new technologies, assists with technology advocacy, and guides the traditional Rehabilitation Engineering Research Center (RERC) staff toward consumer-focused design. The ConLab is designed to be a "think tank" and not a service delivery model. The intent is to develop ideas and create prototypes while leaving direct supply of devices to people in the hands of full-scale manufacturers.
The recruitment of consumers with disabilities is an ongoing process of inviting participation through consumer organizations. A small group of people with physical disabilities has come forward and these volunteers are being provided temporary part-time employment with the RERC. Along with the traditional staff, they are exploring new methods of integrating consumer knowledge and expertise that facilitate a more hands-on approach at the earliest stages of research and development.
PROGRESS--At present 24 consumers in 3 groups are attending 1 to 1 1/2 hour meetings at the duPont Institute on a bi-weekly basis. A standardized questionnaire for moving through the research process has been developed. As their major project focuses, these groups have chosen three devices, an egg-breaking device for use in food preparation, a remotely controlled vacuum cleaner, and a page turner.
All three devices are in various stages of development. A working prototype of the Auto-Vac is being home tested by the members of the ConLab at this writing. The prototype of the page turner is being mocked up and should be in final operation soon. Ideas for the final design of the egg breaker are still being discussed.
FUTURE PLANS/IMPLICATIONS--This project will provide our researchers at with direct input from the consumers about the devices and systems that they are trying to design. In addition, it will provide new sources of ideas and problems.
[145] ASSESSING INDIVIDUALS' PREDISPOSITIONS TO THE USE, AVOIDANCE, OR ABANDONMENT OF ASSISTIVE TECHNOLOGIES
Marcia J. Scherer, PhD, CRC
Center for Assistive Technology, University at Buffalo,
Buffalo, NY 14214; email: mjserd@acsu.buffalo.edu
Sponsor: National Science Foundation, Ethics and Values in Science and Technology, and Program for Persons with Disabilities, Arlington VA 22230
PURPOSE--Past research has categorized predispositions to the use of assistive technologies (AT) as depending upon characteristics within four major areas: a) the particular technology (e.g., design, service delivery), b) the capabilities and personality of the person (e.g., aptitudes, outlook, expectations), c) the nature of the disability (e.g., type, severity), and d) the psychosocial environment of the person (e.g., support from family and friends, life experiences, education). When variables within each of the above areas are organized by category of technology use (optimal and partial/reluctant) and non-use (avoidance and abandonment), individuals can be profiled according to the likelihood of a good match with a particular AT.
METHODOLOGY--The Assistive Technology Device Predisposition Assessment (ATD PA) is a consumer self-report checklist with items of varied format, including 5-point Likert scales. Its purpose is to identify potential sources of person and technology mismatches for early intervention. The ATD PA has subscales to separately assess characteristics of the AT, the temperament of the individual, and the environment in which he or she will use the AT. Side One of the consumer form consists of questions given per consumer on temperament, psychosocial resources, and inquires into individuals' subjective satisfaction with current functioning in many areas and where the person wants the most improvement to occur. Side Two contains 10 questions for consumers to complete per technology on their views of and expectations for that particular AT. Companion professional forms are similarly constructed and allow the assessment of shared perspectives between consumer and professional.
The ATD PA is one set of assessment instruments in the Matching Person and Technology (MPT) Model. Other instruments are the Survey of Technology Use (SOTU) and the Educational Technology Predisposition Assessment (ET PA).
PROGRESS--The ATD PA, SOTU, and ET PA have been administered to users of technologies from childhood to advanced age and in environments ranging from educational settings to acute rehabilitation.
RESULTS--The results of several studies confirm the importance of consumer perceptions and psychosocial factors on decisions to use (or not use) a technology. Characteristics of the technology under consideration further determines its appeal, usability, and utilization.
Specifically, individuals precategorized into five groups according to level of hearing loss were asked to complete the ATD PA and two additional instruments assessing hearing limitations in various environments. The ATD PA was the best predictor of group membership, correctly classifying 85 percent of the participants and providing psychosocial markers associated with awareness of and adaptation to hearing loss.
Another study looked at successful educational outcomes of instruction delivered via telecommunications technologies. One hundred twenty students completed the SOTU, ET PA, the Tennessee Self-Concept Scale (TSCS), and Canfield's Learning Styles Inventory (LSI). The SOTU and ET PA were the best predictors of a) success in the course and b) proficiency/satisfaction with use of the course technologies. Additionally, the SOTU and ET PA were found to provide an efficient and inexpensive alternate means of measuring self-concept and learning style preferences as measured by the TSCS and LSI.
FUTURE PLANS--We are about to undertake a pilot study regarding consumer-directed outcome research on perceived needs, AT use, functional gain over time, and consumer rated quality of life attainment. It is hoped that this information will help lead to better person-technology matching and enhanced consumer AT use and training for use.
[146] DEVELOPMENT OF AN ADAPTIVE TOILETING SYSTEM FOR YOUNG CHILDREN
Steve Ryan, BESc, PEng; Dianna Lee, MA, OT(C); Jan
Polgar, PhD, OT(C); Wes From, MASc; Gloria Leibel, BScPT;
Michael Doell, OCA
Bloorview MacMillan Centre, Toronto, Ontario, Canada M4G
1R8; Department of Occupational Therapy, University of
Western Ontario, Elborn College, London, Ontario, Canada,
N6G 1h7; email: ortcsr@oise.utoronto.ca;
jpolgar@julian.uwo.ca
Sponsor: Ontario Rehabilitation Technology Consortium (funded by the Ontario Ministry of Health)
PURPOSE--An area of concern for parents, daycare providers, and school staff is toileting young children with positioning problems. Unless appropriate postural support is provided on a secure base, this event can be frightening for the young child and disconcerting for the attendant.
Many commercially available devices are inadequate because they are production units that do not offer postural support and are inherently unstable when mounted on a standard toilet seat. A few products do have features that can be adapted to seat the child with a physical disability. However, these devices are cumbersome to store, prohibitive in cost, and do not effectively position the child. This project is designed to identify and incorporate desirable features identified by consumers into a commercially viable product.
PROGRESS--We held focus groups with parents and clinicians to develop consumer criteria and evaluate popular commercial toileting systems. A presentation model was built to illustrate our vision of the product. Parents and clinicians who evaluated it loved it. We subsequently developed a fitting prototype and organized classroom trials to see how well it positioned about 20 children with disabilities. Our findings suggest that the prototype seems to position most children well. This led us to develop a functional prototype for families to use at home. Comments from families who have used it have been encouraging. Links with industry were also cultivated during this time.
FUTURE PLANS--During fiscal 96/97, we will help our industry partner to source appropriate materials and develop a manufacturing and marketing plan. Pre-release testing of the product will be conducted to finalize its commercial design. Product literature and marketing information will be created and tested with consumers.
[147] RAPID PROTOTYPING FOR REHABILITATION AIDS FOR THE PHYSICALLY DISABLED
Richard Mahoney, PhD
Rapid Prototyping Project, Applied Science and
Engineering Laboratories, University of Delaware/A.I.
duPont Institute, Wilmington, DE 19899; email:
mahoney@asel.udel.edu
Sponsor: University of Pennsylvania, Office of Research Administration, Philadelphia, PA 19104; Nemours Foundation, A.I. duPont Institute, Wilmington, DE 19899
PURPOSE--The purpose of this project is to explore the use of modern manufacturing techniques in the design and rapid prototyping of customized rehabilitation aids. In particular, computer-based techniques which integrate computer-aided design, computer-aided manufacturing, and virtual prototyping are being explored.
METHODOLOGY--The goal of this work is to create algorithms and techniques that support the rapid manufacture of customized devices. These techniques apply to three areas: measurement of customer parameters, design of the product, and manufacture of the product. In addition, the process through which a customized rehabilitation aid is designed is being evaluated with respect to traditional engineering design theory.
To evaluate the techniques and algorithms, a number of design case studies are being carried out with new or modified products, including wheelchair trays, feeding aids, ergonomic helmets, and mechanical input devices.
PROGRESS--A demonstration custom manufacturing process
is close to completion. This process includes the following
components:
Interaction with Customers. Project staff have met
with a number of consumer focus groups to determine
important characteristics of both the case study devices and
the process through which measurement of the consumer may
take place. It is the intention to provide consumers with
real devices that may be evaluated, but will also enable
exploration of the manufacturing process.
ProEngineer Detail Design. A detailed design of the
device is prepared on ProEngineer.
Virtual Prototyping. The detail design from
ProEngineer may be ported directly into the Jack human
animation program, which allows the device to be examined in
a virtual environment. Several demonstrations with different
products have shown that this is a powerful tool for
evolving a device design without needing to physically
prototype the device. The simulation software has been
modified to allow a range of interface devices to be
attached, which provides a consumer with the ability to
interact directly with the simulated device. The virtual
simulation also serves to allow for assessment and training
on the prototype device.
ProManufacture. Once the final detail design is
complete, the machining process for the device is specified
in the ProManufacture module of ProEngineer, from which all
tool paths, holding points, tool sizes, and cutting speeds
are specified. The machining commands are directly ported
into a CNC milling machine through a separate
post-processor. This process has been demonstrated with
several test devices and provides a quick and efficient
route between the design and manufacturing process.
Manufacture. Depending on the type of device, it is
manufactured either directly on the CNC milling machine, or
the mill is used to create a mold, which in turn is used on
a vacuum forming machine.
This entire process has been demonstrated and is now in a stage where it will be refined.
FUTURE PLANS/IMPLICATIONS--Immediate plans for this project are to continue to refine the above described manufacturing process and to determine the critical points in the process that affect the costs, in terms of efficiency, time, material, and equipment. Examination with respect to the engineering design process to date has revealed a need to better understand the consumer's role in specifying not only the functional requirements of the product or device, but also the methods through which the customized features are provided.
A further outcome of this project will be the identification of critical technologies that enable cost-effective rapid manufacturing capabilities to be more broadly available.
[148] SPECIAL PROJECTS AND DEMONSTRATION: APPLICATIONS OF TECHNOLOGY TO ENHANCE QUALITY OF LIFE--A COMMUNITY MODEL
Ian R. Pumpian; Andrew Y.J. Szeto; Caren L. Sax
San Diego State University, Interwork Institute, San
Diego, CA 92182-5313; web site:
http://www.interwork.sdsu.edu/projects/
Sponsor: U.S. Department of Education, Department of Special Education and Rehabilitation, Washington, DC 20202
PURPOSE--The project focused on demonstrating that the education and rehabilitation planning processes for individuals with significant disabilities can be enhanced by appropriate applications of technology. The project was designed to demonstrate that persons with severe disabilities using customized adaptations (assistive devices) can participate more meaningfully in integrated work, school, and other community settings; we sought to involve educators, rehabilitation counselors, case managers, teachers, parents, employers, future engineers, and community members with technical expertise in the development of adaptations; and worked to develop a replicable approach for enhancing the applications of assistive technology through direct service, information collection and dissemination, and referrals.
PROGRESS--The above goals were realized through several avenues. To identify individuals who could benefit from customized technical adaptations, the project staff worked closely with teachers and resource specialists of the San Diego Unified School District as well as with staff from local supported employment and supported living agencies. The project targeted transition-aged students (18-22 years) and young adults who, with suitable individualized assistive technology, could become more active participants in school, work, and community settings. Resources were utilized to build up the technical capabilities of local schools and service agencies, expanding the network of rehabilitation and assistive technology professionals, and reducing possible duplication. The project successfully exceeded its goals in terms of number of persons helped and technical adaptations completed.
One key project component of the project was the use of multidisciplinary Tech Teams, individually focused on the specific needs of the consumer. The Tech Teams included friends, family members, interested volunteers, and employers in addition to the special educators, engineering students, OTs/PTs, speech therapists, and community-based rehabilitation professionals who were enrolled in a special seminar jointly taught by the Departments of Special Education and Electrical Engineering. In addition to providing valuable hands-on experience of designing and fabricating a customized assistive device, the seminar facilitated exchanges of ideas and diverse viewpoints. The Interwork Technology Mini-center coordinated various demonstration, training, research, and dissemination activities associated with the project while also serving as a repository of numerous reference materials available for use by the community and Tech Teams. Information about on-going projects and completed projects have been disseminated in print and via the Internet at http://www.interwork.sdsu.edu/projects/.
RESULTS--Major accomplishments of the 36-month project include: 1) better collaboration with the school district, supported employment, and supported living agencies through better utilization of each others' assistive technology resources and expertise; 2) expanded key intra-state and inter-state linkages; 3) design, fabrication, and delivery of over 40 customized technical adaptations; 4) documentation of individualized adaptations using photographs, videotape, technical drawings, and case study descriptions; 5) presentations at the local, state, and national levels; 6) evaluation of the completed projects; and 7) establishment and maintenance of a local web site that describes project activities and accomplishments.
Examples of recently completed adaptations include: A stamping adaptation that ensured accurate rubber stamp imprints on student passes modified for use in a counseling office by a high school student with cerebral palsy. Plywood foot guides attached to Nordic Track;tm exerciser for use by a man with hemiparesis. Padded arm guides attached to power wheelchair for a woman with cerebral palsy. The adaptation held a water bottle in place while also enabling easier passage through narrow doorways. A head control switch used by a man with head injury modified to keep his head in an upright position. A plastic transfer device designed to enable a transition-aged high school student to fill salt and pepper shakers at a restaurant. The door of a van modified for easier opening and closing by a man with paraplegia. A flat bed cart (or dolly) modified for use by a man with cerebral palsy using a manual wheelchair in his job at a nursery. A custom harness fabricated and a trumpet modified (by adding levers attached to its keys) so that a high school student who has a spinal cord injury could once again play that musical instrument. A ticket tearing device modified for use by a man with cerebral palsy at movie theater job. Flexible and inexpensive electronic communication devices designed and built to meet the rudimentary communication needs of students with significant disabilities in general education classes. A lazy susan to hold files and papers for a man with quadriplegia; also designed custom speaker mounts for his augmentative communication device.
[149] TRANS-TRAIN: TRANSDISCIPLINARY TRAINING OF REHABILITATION PERSONNEL IN ASSISTIVE TECHNOLOGY
Ian R. Pumpian; Andrew Y.J. Szeto; Caren L. Sax
San Diego State University, Interwork Institute, San
Diego, CA 92182-5313; web site:
http://www.interwork.sdsu.edu/projects/
Sponsor: U.S. Department of Education, Department of Special Education and Rehabilitation, Washington, DC 20202
PURPOSE--TRANS-TRAIN is a 36-month project that seeks to provide preservice and inservice training to rehabilitation personnel in Assistive Technology (AT). It is a university-based program that combines academic classroom instruction with experiential field activities. Although discipline specific training will be given, TRANS-TRAIN fundamentally is a transdisciplinary project that establishes a series of courses, guided design projects, and internships that focus on the development and use of AT. To complement an existing College of Education certificate program in "Supported Employment and Transition," a specialization area in "Rehabilitation Technology" is being developed for the Department of Electrical and Computer Engineering.
METHODOLOGY--Because the incoming students come from various educational and vocational backgrounds (e.g., engineering, special education, rehabilitation counseling, and communicative disorders), the certificate program can be customized to fit their backgrounds, skills, interests, and intended application areas. In addition to the six to nine unit curricula (supported in part by project funds), students seeking a "Certificate in Assistive Technology" complete six units of formalized discipline-specific course work from within their home departments and three to six units of transdisciplinary seminars covering a broad range of rehabilitation technology competencies and knowledge. For hands-on experience, students participate in a number of internships, off-campus and on-campus, under the supervision of professors and practicing professionals in rehabilitation engineering, special education and rehabilitation, and communicative disorders.
One key component for training personnel in the development of customized assistive adaptations is the use of transdisciplinary Tech Teams individually organized according to the specific needs of an individual with disabilities. These Tech Teams encompass friends, family members, employers, and volunteers, special educators, engineering students, OTs/PTs, speech therapists, and community-based rehabilitation professionals enrolled in TRANS-TRAIN sponsored classes. In addition to providing valuable hands-on experience of designing and fabricating a customized assistive device, the classes foster exchanges of ideas and viewpoints among persons from varied backgrounds. To monitor progress and insure that the assistive technology is consumer-driven and integrated into the planning processes, a series of 12 milestones (e.g., a request for assistance, research and data collection, design, prototype construction, field-testing, and evaluation) is used. To promote dissemination and replication by others, customized adaptations designed and constructed through this project are being incorporated into a database that will track information about each adaptation.
PROGRESS--Between fall 1995 and spring 1996, 12 undergraduate and 2 graduate engineering students, 18 graduate students in Special Education, and 13 graduate students in rehabilitation counseling participated in the transdisciplinary seminar.
Tech Teams were formed around individuals with disabilities who needed some sort of adaptations. Some of the adaptations designed, fabricated, field tested, and delivered include:
FUTURE PLANS--During its third year, TRANS-TRAIN will seek formal recognition and approval of a certificate program in AT from the Department of Electrical and Computer Engineering, the College of Engineering, and the University. TRANS-TRAIN will continue to offer an advanced version of the transdisciplinary seminar in AT, set up initial internships at local agencies with appropriate experts, and give a course in engineering on "Electronic Devices for Rehabilitation."
[150] A LOW COST, HORSE-DRAWN CART FOR INDIVIDUALS WITH DISABILITIES
Beth A. Todd, PhD; H. Nikole Brock, BS; S. Leeann Smith,
BS; P. Dawn Stripling, BS
University of Alabama, Tuscaloosa, AL 35487-0276; email:
btodd@coe.eng.ua.edu
Sponsor: None listed
PURPOSE--A low cost horse-drawn driving cart was needed by a therapeutic riding program for individuals whose disabilities would not let them ride a horse comfortably or prevented them from being covered under liability insurance any other way. The cart was designed to permit the individual with disabilities to control the horse with the reins.
METHODOLOGY--The design criteria which were considered included issues related to the design for "generic" disabilities. However, the cart was primarily designed for someone who is confined to a wheelchair and has little upper extremity motion and trunk control. Design criteria related to such issues as wheelchair tie-downs, seating, cart dimensions, tires, braking system, and suspension system. While this cart will be moving at much smaller velocities than a mechanically powered vehicle, information on transportation of wheelchairs, such as that being developed by the Subcommittee on Wheelchairs and Transportation (SOWHAT) was included in the design.
PROGRESS--The cart has been designed and built. It is currently under evaluation by a therapeutic riding center.
RESULTS--The cart was built from wood and steel with a materials cost of approximately $400. A fifth wheel was attached to the top of the front axle to allow it to rotate independently of the cart body. A ratchet strapping system was used to tie the wheelchair of the driver down to a piece of angle iron on the floor of the cart. Trailer tires were used to keep the cart low to the ground and provide rough terrain capabilities. The braking system is a locking system for loading and unloading the driver on a ramp in the rear of the cart. The cart does not have a dynamic braking system.
[151] ASSISTIVE ROBOTICS IN A VOCATIONAL SETTING
Richard Mahoney, PhD; Jill Schuyler; Patty Hove, OTR;
Mike Meyreles
RERC on Rehabilitation Robotics, Applied Science and
Engineering Laboratories, University of Delaware/A.I. duPont
Institute, Wilmington, DE 19899; email:
mahoney@udel.edu;
schuyler@asel.udel.edu;
smartcen@ravenet.com
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 20202; Nemours Foundation, A.I. duPont Institute, Wilmington, DE 19899
PURPOSE--The purpose of the Vocational Robotics project is to study the issues involved in the employment of individuals with manipulation disabilities through the use of interactive robotic devices. The goals of this project are two-fold. First, this project aims to employ individuals with disabilities in real mainstream jobs, through the use of assistive robotics. This will be accomplished through surveying a wide range of vocations and evaluating the manipulation requirements of their job tasks. Also, a collaboration with a local rehabilitation center that is actively involved in job placement of people with disabilities will provide a site for assessment and training of potential employees. Second, this project intends to identify the process through which a suitable robotic system may be designed so that a potential job may be made accessible to a person with a manipulation disability. This task involves identifying the organization that addresses both engineering and non-engineering issues related to the potential employees disability. It is expected that a successful job placement will require input from an engineering design team, vocational rehabilitation specialists, representatives from the potential employer, and funding sources.
METHODOLOGY--Three areas of concern are being addressed in this work: assessment and training, system integration, and job identification and analysis. The ultimate goal of the project is to stimulate the development of areas of expertise within normal vocational rehabilitation establishments. This expertise must extend to job site evaluations appropriate for implementation of robotic systems, knowledge of existing commercial robotic technology or the ability to design and construct suitable one-off systems, knowledge of available user interface approaches and software, and facilities for assessing the aptitude of a potential employee not only to work with a robot but also to carry out preliminary training. It is hoped that through this project, the appropriate application of robots in vocational settings will grow to the point where it is a viable and acceptable alternative.
PROGRESS--We have completed 3 years of the vocational robotics project. In the first year, we focused on identifying individuals who work in the vocational assessment field. An occupational therapist trained for a year on the robot and designed several assessment/evaluation tasks. During the second year, we furthered the training for the occupational therapist as well as demonstrating the use of many different interface devices to operate the robot. A demonstration robotic workstation was installed in the Easter Seal rehabilitation facility to assess individuals as well as to be seen by many persons who may be in need of a manipulator. A data acquisition routine for Cambridge University Robot Language (CURL) use was customized to obtain the maximum amount of information about its use at Easter Seals. In addition to the assessment aspect of the training, a couple of different routes were explored to find particular types of jobs for individuals with a robotic workstation. First of all, the Rehabilitation Services Administration database of the clients of all of the Vocational Rehabilitation Centers was analyzed as to the characteristics of the individuals who did and did not receive jobs. Also, the Dictionary of Occupational Titles was explored using the Valpar System 2000 software to find out what types of jobs an individual with disabilities and a robotic workstation could find. Finally, individual jobs in the surrounding area were explored to find their particular responsibilities to see if and how a robotic adaptation could help.
Our goal in the third year has been to gain exposure for the assitive robot and to receive feedback on its development. A morning seminar was held to inform service providers and a group of people from various vocational evaluation centers about the system and get their responses.
FUTURE PLANS/IMPLICATIONS--The primary objective is to provide an appropriate robotic accommodation in a real work environment. This accommodation will provide expertise and exposure that will become the foundation for future robotic accommodations.
[152] A BODY-POWERED REHABILITATION ROBOT
Tariq Rahman, PhD; Sean Stroud, MS
RERC on Rehabilitation Robotics, Applied Science and
Engineering Laboratories, University of Delaware/A.I. duPont
Institute, Wilmington, DE 19899; email:
rahman@asel.udel.edu
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 20202; Nemours Foundation, A.I. duPont Institute, Wilmington, DE 19899
PURPOSE--The goal of this project is to develop a technologically simple, wheelchair-mounted manipulator to allow a person with no or very little arm function to interact with his surroundings. The robot will be controlled and at least partially powered through bowden cables by the intact motion capabilities of the user, such as the head or hand. This method of robot control relies on extended physiological proprioception (EPP), the same control principle used in cable-driven prostheses. It has been shown that a system which compliments visual feedback with sensory channels is superior to visual feedback alone. The intended population that would benefit from such a device has physical disabilities such as spinal cord injury, multiple sclerosis, and cerebral palsy.
METHODOLOGY--The following design specifications drive
the methodology of the project:
Intuitive and easy to use. The inputs of the user
should map in an integrated manner to the outputs of the
manipulator: a proportional, three-dimensional position
mapping of the user's input position signal to the position
of the arm's gripper is desired. A direct connection between
the user interface and arm facilitates a system which is
easy to use, since proprioception and force reflection are
naturally built into the control system.
Modular. The system will be modular in two senses.
First, the arm will accept several different user inputs.
These inputs depend upon the available user body motions,
which to a large extent, depend upon the user's disability.
For example, if the best available user input is from the
head, the arm needs to accommodate whatever interface is
designed for head input. Another way the system needs to be
modular is in its ability to accept power assist units. In
this way, if the user cannot supply sufficient power to the
interface to directly cause the arm to move, power amplifier
modules will be added to specific joints to assist the user
in operating the arm. The issue of desiring minimal user
input power naturally leads to the requirement of arm
gravity compensation throughout its range of motion.
Cost. The high cost/usefulness ratio of most
rehabilitation robots makes their use very limited. It is
the goal of this project to maintain a simple design
philosophy so costs can be kept to a minimum.
Aesthetics. The arm is designed to geometrically and
functionally resemble a human arm. The interface unit will
be designed to be as unobtrusive as possible, and the cable
routing will be neat.
PROGRESS--Two arm prototypes and two interface units have been designed and constructed. The second has been tested with both interface units, one for the head and the other for the hand. One of the design objectives was to have an end-point controlled, mechanical linkage which resembles the human arm. To facilitate this objective, a spherical coordinate system was chosen for the arm with an extra degree of freedom added to kinematically couple head input to arm motion. A direct mapping exists between the yaw, pitch, and roll axes of the arm and head interface, while a proportional mapping is present between the linear, horizontal motion of the user's head and the radial motion of the arm. Low friction bowden cables connect the arm to the head interface unit. A four bar linkage design of the arm's main beams is to allow gravity compensation of the mechanism throughout its full range of vertical motion.
FUTURE PLANS/IMPLICATIONS--It was decided through in-house evaluation to build a third generation arm and a second generation head interface unit. In the present system, high friction makes it difficult for the user to rotate the arm along some axes. A different gravity balancing technique will be used to decrease the friction throughout the system and a minor design constraint has been relaxed, thus making the system significantly simpler. Some additional issues to be addressed include attaching a gripper to the arm, donning/doffing, adjustability to accommodate different users, further development of power assist units, and exploring methods of allowing the user to lock the system.
[153] REHABILITATION ROBOTICS INFORMATION PROGRAM
Julia Mercier, BA; Richard Mahoney, PhD; Richard Foulds,
PhD
RERC on Rehabilitation Robotics, Applied Science and
Engineering Laboratories, University of Delaware/A.I. duPont
Institute, Wilmington, DE 19899; email:
foulds@asel.udel.edu;
mercier@asel.udel.edu
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 20202; Nemours Foundation, A.I. duPont Institute, Wilmington, DE 19899
PURPOSE--The Information Program in Rehabilitation Robotics collects, organizes, synthesizes, and disseminates information on rehabilitation robotics to professionals, consumers and their families, manufacturers, and other researchers.
METHODOLOGY--The information program produces technical reports, book chapters, journal articles, and conference and symposium presentations that describe the research, development, and evaluation work underway at ASEL. The project also produces a number of resource materials and activities, including: videotapes, a newsletter, a web site, conferences, and workshops on robotics topics.
PROGRESS--Robotics research staff have produced 22 publications that have been disseminated through the Robotics Information Program. Inquiries for informational materials and technical assistance have been personally answered. A Rehabilitation Robotics Research Program web site has been developed and refined. Two videotapes on robotic products and laboratories have been completed. The Rehabilitation Robotics Newsletter continues to be produced and disseminated. The program established the Consumer Innovation Laboratory, which publishes a quarterly newsletter, maintains a web site, and holds bi-weekly design meetings.
FUTURE PLANS--The RERC will hold a workshop on design, control, and implementation of wheelchair-based manipulators. The robotics web site will continue to be expanded and updated. A series of videotapes on specific robotics projects are planned. Research progress and findings will be more widely disseminated through a redesigned information program.
[154] IMPROVING THE FUNCTIONAL UTILITY OF REHABILITATION ROBOTICS THROUGH ENHANCED SENSORY FEEDBACK: THE VIRTUAL HEADSTICK
Tariq Rahman, PhD; Richard Mahoney, PhD; Edwin A.
Heredia, PhD; Shoupu Chen, PhD; Randy Glass, BS
RERC on Rehabilitation Robotics, Applied Science and
Engineering Laboratories, University of Delaware/A.I. duPont
Institute, Wilmington, DE 19899; email:
rahman@asel.udel.edu
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 20202; Nemours Foundation, A.I. duPont Institute, Wilmington, DE 19899
PURPOSE--A major advantage of mouth sticks and head sticks as extension devices for people with disabilities is the extended proprioception. Proprioception is the ability to sense forces and other perceptual cues present at the tip of the extension device. The conventional mouth stick, for example, can successfully be used in daily activities due to its good proprioceptive properties: it provides an intimate contact of the device with the user's mouth which is rich in tactile and proprioceptive sensing ability, and it is lightweight and very stiff, and therefore conveys tactile and kinesthetic information from the environment. However, there are two important limitations for this type of device. The first is the spatial limitation caused by the fixed length of the mouth stick plus the restricted motion of the human head. The second is the amount of mechanical power that can be transmitted to the end of the mouth stick. The longer the mouth stick, the more the power required to achieve similar tasks.
We are developing a robotic teleoperation system that, in principle, can overcome both limitations of conventional mouth sticks. Teleoperated robots can deal efficiently with the problem of spatial limitation, and since power is generated at the robot-arm itself, the second limitation can also be alleviated. However, proprioception and the exchange of kinesthetic information using teleoperation robots are issues that need to be carefully examined.
METHODOLOGY--We have developed a teleoperation system based on two subsystems: a head controller (master) and a robotic arm (the slave). Head motion is sensed at the controller end and transmitted digitally to the robotic arm, which attempts to track the original head motion. The amplification and filtering of motion and forces can be easily implemented at the information processing stage. The master controller is a six-degree-of-freedom (DOF) manipulator, originally designed for hand control, and modified for head operations using an attached helmet. The slave robot is a Zebra Zero robot with six DOF, lightweight, and initially designed for simple industrial applications. Forces and positions at the slave end can be measured, digitized, and sent back to the master-controller computer. The computer uses this data to produce forces in the helmet that mimic the ones encountered by the robot arm, providing a certain degree of proprioception.
PROGRESS--This force-reflecting teleoperation system has been successfully implemented. While no accidents have resulted from testing this equipment, the safety issue has been an important concern during the last year, mainly because the ultimate users will be human beings. Mechanical safety devices have been investigated, and electrical safety devices and monitoring software have also been implemented to prevent the system from reaching high power values.
Besides safety, different teleoperation architectures have been implemented: 1) a Cartesian-based control where head motion is mapped to robot-arm motion, 2) a rate control system where head motion is mapped to robot-arm velocities, and 3) a spherical control unit where rotational head motion is prioritized and mapped to robot-arm motion.
Finally, a head stick unit has been designed and developed with the purpose of comparing the performance of teleoperation systems with conventional head sticks. Different tasks related to daily life activities have been established for performance evaluation. The tasks attempt to measure proprioception ability, power transference capacity, space limitations, and the conveyance of kinesthetic information.
FUTURE PLANS--We are in the process of evaluating the performance of the different architectures for the head-controlled teleoperation system. Our immediate goal is to establish a series of comparative results that may be used as the basis of future developments in this area. Future issues to examine include: 1) the design and development of full-bilateral architectures where forces and positions are transmitted from the master to the slave and vice versa, 2) the mounting of a similar structure to a wheelchair, and 3) power control and amplification to compensate for motion limitations of the head.
[155] MULTI-MODAL CONTROL OF A REHABILITATION ROBOT
Richard Foulds, PhD; Shoupu Chen, PhD; Daniel Chester,
PhD; Zunaid Kazi, MS; Matthew Beitler, BS
RERC on Rehabilitation Robotics, Applied Science and
Engineering Laboratories, University of Delaware/A.I.
duPont Institute, Wilmington, DE 19899; email:
foulds@asel.udel.edu
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 20202; Nemours Foundation, A.I. duPont Institute, Wilmington, DE 19899
PURPOSE--The rehabilitation robotics research literature describes many demonstrations of the use of robotic devices by individuals with disabilities. In general, the existing interface strategies have not met the desires of the disabled community. The conventional prototype interfaces have taken two approaches to achieving effective use by individuals with disabilities. Many have commands which are issued by the user and activate the robot to perform preprogrammed tasks. Others have sought to give the user direct control of the manipulator. In this project, a new hybrid interface strategy is designed. This new man-machine interface combines command and control approaches to provide for user direction of the robot through the use of multiple modes of interface in conjunction with sophisticated capabilities of the machine. Users of this system use gestures (pointing) to indicate locations, and spoken commands to identify objects and actions. The use of multiple modes of control and command allows the user to operate the robot in a manner which more closely matches the needs of the user. The operation is expected to be superior to conventional methods since it capitalizes on the strengths of the user's abilities and coordinates these abilities with software and hardware sophistication of the robot and computer technology.
METHODOLOGY--This multi-modal approach is based on the assumption that the world of the user is unstructured, but that objects within that world are reasonably predictable. There are two major components of this hybrid interface strategy, including a system that determines the three-dimensional (3-D) contours of objects and surfaces which are in the immediate environment, and an object-oriented knowledge base and planning system which superimposes information about common objects in the 3-D world. The effectiveness of this approach can be demonstrated in the following example. An individual with a disability uses an electric wheelchair and a portable robot arm. He wishes to move the pen, which is on the desk, to the box: using a head laser pointer, he points to the pen and says, "move." The user then points to the box, and says, "there." The combination of the initial pointing accompanied by the command tells the robot to pick up an object at a specific location. The combination of the subsequent pointing and command tells the robot where to move the object.
PROGRESS--Integration. The different
developmental subsystems have been integrated into a working
system which is now undergoing in house-testing. The
components RoboMind (the intelligent robot planner), RoboArm
(the robot arm control mechanism), RoboEye (the vision
system), and RoboEar (the speech recognition mechanism) have
now been integrated into a working system by means of RPC
connections.
HCI subsystem. Significant progress has been
achieved in the user interface portion of the system: 1) The
communication protocol between the user and the system has
been completed. 2) A grammar has been developed to codify
the human-computer communication protocol. 3) A parser based
on this grammar has been constructed which understands a
subset of natural language-like instructions (e.g., move the
blue book to the right of the pen).
Planning subsystem. The knowledge-based planning
mechanism is nearing completion. Multi-modal instructions
combining speech and gesture are parsed and interpreted by
the planner. The planner understands a variety of low level
robot arm manipulatory instructions as well as a number of
high level pick and place instructions. Work is progressing
in allowing users to teach the system new high-level tasks
for later re-use and/or adaptation. A basic knowledge base
of objects has been constructed that allows researchers to
test planning activities on a variety of everyday objects.
Partial plan-monitoring, error detection, and recovery
techniques have been constructed and will be completed
during this year. A plan adaptation mechanism has been
completed that allows the planner to adapt previously used
plans to synthesize new plans. A plan learning mechanism has
been implemented which allows the user to "teach" the
robot to learn new tasks.
Vision subsystem. The performance of the 3-D vision
system has been improved significantly in terms of speed and
versatility; we have 1) incorporated the SGI Media Video
Library routines to perform automatic video channel
alternating for real-time image capturing; 2) modified the
laser spot location searching strategy to reduce the
locating time to less than 1 second; 3) implemented a color
vision module for better object-disambiguation for the
planner system; 4) completed a contour-based stereo-matching
method for rapidly (in the range of 15 to 30 sec, depending
on the number of objects involved) recovering 3-D
information for circular and linear image components; 5)
developed a Hough transformation based 3-D information
segmentation for providing the planner with the object pose
parameters; and 6) ported the structured-light stereo
matcher from the Visilog environment to C for faster
performance and faster user interaction cycle.
Simulation subsystem. The graphic display system has
been completed and is soon to be undergoing user testing.
The display mechanism has also been partially integrated
with the planning mechanism.
FUTURE PLANS--Fine tuning and testing with consumers.
[156] DEVELOPING A ROBOTICALLY AIDED SCIENCE EDUCATION LABORATORY FOR STUDENTS WITH SEVERE PHYSICAL DISABILITIES
Richard Howell, PhD; Brian Chipman; Jill
Schuyler
RERC on Rehabilitation Robotics, Applied Science and
Engineering Laboratories, University of Delaware/A.I. duPont
Institute, Wilmington, DE 19899; email:
rhowell@magnus.acs.ohio-state.edu
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 20202; Nemours Foundation, A.I. duPont Institute, Wilmington, DE 19899
PURPOSE--The intention of this project is to develop a functional, stand-alone, educational robotic system for children with severe physical disabilities. This system is to be complete with a prototype robot, assessment and educational curriculum materials, and other supportive documents. The goals of the project include prototype development, therapeutic assessment, training and education, and dissemination. Pilot implementations to investigate the feasibility of the system in a series of field tests are under way in classroom settings in the Brandywine Public Schools and the Columbus Public Schools by the Applied Sciences and Engineering Laboratories (ASEL) and The Ohio State University (OSU), respectively. The proposed prototype research and development project expands upon an extant foundation, providing for the eventual integration of the science laboratory, accessible instruments, software tools and robotic manipulation abilities into a complete science laboratory environment. This setting will someday enable young learners who have severe physical disabilities to work with greater independence within a powerful laboratory-based setting, incorporating the best tools and instructional strategies available.
METHODOLOGY--A mixed methodological approach that integrates quasi-experimental and qualitative methodologies will be used to gather data on the academic performance and cognitive, psychomotor, and affective impact of using the prototype laboratory environment. One of the primary functions of this project is to develop a science educational curriculum which is field tested and validated. The framework of the educational curriculum involves a two-phase process in which students are first trained to use the robot hardware and software using simple object manipulation activities. The second phase involves the development of science education activities which follow the specified sequence of: explore, observe, think, find out, and record. Each of these areas are used to develop a contextualized understanding of the scientific phenomenon that are under investigation in addition to doing the hands-on experimentation using the robotic system. The students include both disabled and nondisabled students who will be working in research teams at the field site in the schools.
PROGRESS--Interface Development. Development
of a software interface that is able to engage the students
at the appropriate level, as well as provide sufficient
flexibility for adaptations by teachers, continues.
Development time has been spent investigating an approach
using Labview software rather than the commercial version
of the Cambridge University Robot Language (CURL) because of
inadequacies in CURL.
Site Evaluation with Students. Experimental
evaluation of student use of the system has commenced in the
field site school at OSU. Preliminary indications are that,
within the teams of both disabled and nondisabled students,
the robot-based science activities support and emphasize
science learning.
ASEL field site. Progress has been made in
evaluating various interface alternatives, such as
DragonDictate and Intellikeys.
RESULTS--The first stage of field testing at OSU will be completed. The results of this study will be used to specify more exactly a science-based curriculum based on the robot. Further efforts will be made to package the findings into a curriculum that may easily be transferred to other sites, and that will eventually form the basis of a commercial product.
[157] CONTROL AND SIGNAL PROCESSING STRATEGIES FOR TREMOR SUPPRESSION
Kenneth E. Barner, PhD; Tariq Rahman, PhD; Edwin A.
Heredia, PhD; Qing Xu, MS
RERC on Rehabilitation Robotics, Applied Science and
Engineering Laboratories, University of Delaware/A.I. duPont
Institute, Wilmington, DE 19899; email:
barner@asel.udel.edu
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 20202; Nemours Foundation, A.I. duPont Institute, Wilmington, DE 19899
PURPOSE--Tremor is a rhythmic uncontrollable oscillation that appears superimposed to voluntary movement. Approximately 0.4 percent of the population in the U.S. is affected by some kind of pathological tremor. The amplitude of the tremor distortion can be high enough to disable the subject in normal daily operations such as eating, writing, computer operation, and wheelchair operation. This project is designed to allow people with pathological tremor to have normal manipulation in human-computer interaction using hand-controlled input devices such as joysticks, mice, and telemanipulators.
METHODOLOGY--Three approaches have been developed for tremor suppression, two of which are signal processing approaches and the other is force feedback approach. In each case, a subject is asked to perform either a pursuit tracking task or a writing test. A PerForce hand controller, a force reflecting device originally designed for the space station, is used as the input device. In the signal processing approaches, the hand controller is used as a simple unpowered input device and digital filters are applied to the hand/arm motion signal collected by the computer. In the force feedback approach, force is generated by the motors built in the hand controller and applied on the handle to cancel out the tremor in the arm motion. The subject would feel resistance from the controller, while in the signal processing case, the subject could only see the effect of the filter on the computer screen.
PROGRESS--The first signal processing approach is tremor equalizers. Equalization is a technique that has been successfully used over the years in the field of telecommunications. However, several important contributions, such as Filtered Mean-Square Error with delay correction, have been achieved to make it appropriate for tremor cancellation. Based on the data collected from the pursuit tracking tasks, optimal equalizers are developed to cancel exactly the tremor frequency bands and leave the intention movement as intact as possible. Pulled optimization technique is used and it brings an elegant way to design the equalizer delay.
The second signal processing approach is an adaptive finite impulse response (FIR) filter. The adaptive FIR filter is able to learn about the tremor signal and eventually to produce a signal that cancels it. This approach can handle tremor with multiple frequencies and with slowly time-varying characteristics. Because it is parallel to the main signal flow, the filter introduces no time delay to the voluntary movement signal. Another advantage of the adaptive filter is that no information other than the frequency range of tremor has to be known beforehand.
In the force feedback control, the damping force has been explored and tested. The effect of applying damping force is equivalent to increasing the damping of the mechanical system. This is called active damping and is more flexible than the usual mechanical damping. Tremor can be significantly reduced by increasing the damping force. One important accomplishment in this approach is system modeling, which is the basis of controller design.
Besides, some efforts have been made to explore the possibility of replacing PerForce with a new sensible device, Phantom. PerForce has proved to have several hardware restrictions on implementing more effective force controllers. Damping force has been implemented on Phantom and it has been proved that sufficient force can be provided to suppress tremor.
FUTURE PLANS/IMPLICATIONS--A more complicated controller will be designed to have more effective force feedback control on tremor. Stability will remain a big issue and needs further investigation. The working environment will be changed from DOS to Microsoft Visual C++ under Windows NT workstation.
[158] AUTOMATIC MODE SELECTION IN A SHARED CONTROL SYSTEM
Simon P. Levine, PhD; Richard C. Simpson, MS
Department of Biomedical Engineering, Rehabilitation
Engineering Program, University of Michigan Medical Center,
Ann Arbor, MI 48109-0032; email: silevine@umich.edu,
rsimpson@umich.edu
Sponsor: National Institutes of Health, National Center on Medical Rehabilitation and Research Bethesda, Maryland 20892
PURPOSE--Shared control systems are defined as those human-machine systems where control is shared between a human operator and the machine components. The NavChair Assistive Wheelchair Navigation System is an example of a shared control system developed to provide improved mobility for people with severe disability. The NavChair shares control decisions with the human operator to reduce the motor and cognitive effort required to operate a power wheelchair. During development of the NavChair it was observed that one set of operating parameters could not provide all of the functionality desired of the chair. For this reason, the NavChair offers several different modes of operation: general obstacle avoidance, door passage assistance, and automatic wall following. The presence of multiple operating modes necessarily creates the need to choose between them. Requiring the wheelchair operator to perform the task of mode selection could place unacceptable performance burdens on a large portion of the target user population. Instead, we are developing and evaluating methods to infer the correct operating mode automatically.
METHODOLOGY--The mode selection method used by the NavChair must meet several design criteria. Obviously, it must make the correct operating mode decision as often as possible. Another important criterion is that it avoid frequent mode changes, which could lead to an uncomfortable ride for the operator. Finally, decisions must be made in real-time, which is not a trivial concern given the NavChair's computer hardware (a 486/33 personal computer) and the need to share processing time between navigation assistance and automatic adaptation.
The approach we have taken is to combine several limited information sources together into one coherent automatic adaptation mechanism. These information sources are combined using probabilistic reasoning techniques developed by the artificial intelligence community.
PROGRESS--Currently, the NavChair makes use of two information sources to make adaptation decisions: (1) its location within its global environment and (2) the identities of objects in its immediate surroundings. The NavChair's location is tracked within an internal map, and each location within the map has its own set of probabilities as to how likely each operating mode is in that location. For example, within a narrow hallway, wall following and door passage mode are more likely than general obstacle avoidance, while the opposite is true in a spacious room. Objects in the wheelchair's proximity are identified from the same sonar signals that the NavChair uses to avoid obstacles. Currently, the NavChair has robust and efficient methods for identifying both walls and doors. During operation, the NavChair constantly updates its beliefs in its location and the presence (or absence) of doors and walls in its surroundings. Every time these values change, they are used to make a new decision regarding the most appropriate mode for the wheelchair. If the decision to change operating modes is made, the operating parameters that define the new mode (what behavior to exhibit, how fast the chair is allowed to travel, how close the chair can approach obstacles) replace the parameters that defined the previous operating mode.
RESULTS--Preliminary results indicate that our approach to automatic adaptation meets the design criteria specified above. Experiments using nondisabled subjects indicate that the NavChair can perform nearly as well adapting on its own as it can when mode selection is controlled by an experimenter. In addition, combining the individual information sources has been shown to produce better performance than either of the individual sources could provide alone.
FUTURE PLANS--Our recent work has focused on incorporating as much information as possible into the mode selection process. In the future, we hope to include even more information sources such as user control models and additional sensors.
[159] A STUDY OF SHOULDER FUNCTION AS AN INPUT TO AN ASSISTIVE ROBOTIC SYSTEM
David Fass, MS; Tariq Rahman, PhD; Rami Seliktar,
PhD
Applied Science and Engineering Laboratories, University
of Delaware/A.I. duPont Institute, Wilmington, DE 19899;
email: fass@asel.udel.edu;
rahman@asel.udel.edu;
seliktar@asel.udel.edu
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 20202; Nemours Foundation, A.I. duPont Institute, Wilmington, DE 19899
PURPOSE--The purpose of the Shoulder Control project is twofold. First, the study attempts to quantify the capacity of the shoulder to provide a control signal for an assistive robotic system based on extended physiological proprioception (EPP). This kind of robotic system (in particular, one having the bilateral master-slave architecture) is considered to have unique features, including position unbeatability and force-feedback, which may greatly improve the quality of interaction between the machine and its human operator. The utility in controlling a robot with the shoulder would be most apparent with regard to individuals with quadriplegia who retain some control of the shoulder in addition to control of the head and neck, in which case the shoulder site could complement the head and neck as a source of control inputs into the assistive system. To achieve the desired quantification of shoulder behavior, the performance of the isolated shoulder is compared to performance of the full upper limb in accomplishing identical control tasks which simulate the kind of control which would be typically required in operation of an EPP-based assistive system.
The second aspect of the study is an assessment of the impact of "control impedance" on performance of both the isolated shoulder and full upper limb. The motivation for this phase of the study is the finding that varying control parameters (i.e., qualities, like damping, which contribute to the "feel" of a control) can impact performance in performing tasks using the control. Because of the heavily software-oriented architecture of the master-slave robotic system, there is a great flexibility in making adjustments of this kind. Therefore, there is an opportunity here to illuminate what impact, if any, tuning the system impedance in various ways will have on performance with an EPP-based system. The information obtained from this second dimension of the study has relevance as well to control originating from other sites, such as the head and neck.
METHODOLOGY--To investigate these aspects of shoulder performance, a variable-impedance test device, the Shoulder Interface Mechanism (SIM), was designed to interface with a subject's right shoulder. Additionally, the SIM can be fitted with a joystick-type handle for tests using the hand and arm. The SIM is a parallel linkage actuated by two electric motors driven by HCTL 1000 motor control chips, and carries a two-axis force sensor above the shoulder harness. In use, the SIM behaves as a passive load coupled to the operator's shoulder (or hand), responding to the force applied by the shoulder (hand) to the harness force sensor in a manner determined by the compliance and damping properties established for it by the microcomputer controller.
Shoulder performance is being assessed through use of tracking tasks that are chosen to be representative of the range of movements the shoulder might be called upon to make as a control source in a manipulation system. During testing, position, velocity, and force data are recorded from the SIM by the computer at a rate of approximately 17 Hz. The data collected will be used to analyze how the two control sites (isolated shoulder and full upper limb) are impacted by system impedance alterations throughout their ranges of motion.
PROGRESS--This work is in its final stage. All relevant data have been acquired and the statistical analysis is being completed. Early results show that tracking performance, especially with the shoulder, is responsive to impedance tailoring.
[160] ADVANCED INFORMATION RETRIEVAL
Marianne Hickey, PhD; Alan F. Newell, PhD
MicroCentre, Applied Computer Studies Division,
Department of Mathematics and Computer Science, University
of Dundee, Park Wynd, Dundee DD1 4HN Scotland, UK; email:
mhickey@mic.dundee.ac.uk
Sponsor: EC Human Capital and Mobility Programme
PURPOSE--Our aim is to develop a highly efficient interface by which a physically disabled nonspeaker can use a computer to communicate reusable conversational material, as well as to evaluate the use of different intelligent, lexicon-based information retrieval techniques for a communication aid.
METHODOLOGY--Users can potentially communicate faster by selecting whole prestored utterances, rather than by constructing every message character by character. However, current systems impose a high cognitive load on the user who must remember access codes. The project will tackle this problem by using advanced text retrieval methods, including automatic morphological analysis and searches for semantically related words.
PROGRESS--A new interface for a Windows 95 environment is being developed. A larger lexicon containing information about semantic relatedness and morphological code has been compiled. Evaluation of the new interface began in autumn 1996.
[161] SIGN PS--THE DEVELOPMENT OF AN INTERACTIVE PRINTING SYSTEM FOR SIGN LANGUAGES
Alistair Y. Cairns, PhD; Ian W. Ricketts, PhD; Peter
Gregor, PhD; Craig D. Ramsay, BSc; Julian F. Cracknell,
BSc
MicroCentre, Applied Computer Studies Division,
Department of Mathematics and Computer Science, University
of Dundee, Park Wynd, Dundee DD1 4HN Scotland, UK; email:
acairns@mic.dundee.ac.uk
Sponsor: EC--Telematics Programme
PURPOSE--Research has shown that most deaf people do not reach functional literacy. Most communication within this group occurs by means of sign languages, but currently there is no conventional written form of these languages. The aim of the SIGN PS project is to develop a system which will support the production of documents which have been created with, and are written in, sign language.
METHODOLOGY--This will involve investigation into the computer recognition of sign language using specialized methods, such as instrumental gloves and video images; the development of an attractive sign font to graphically represent sign language in printed format; the development of a predictive mechanism to reduce user effort in creating documents and to improve the sign recognition accuracy of the system.
PROGRESS--A font for representing signs had been developed, in conjunction with software to allow the user to manipulate the font elements. This consists of a sign editor, which is used to group the font elements into an individual sign, and a preliminary document editor which is used to form the signs into a document written in Sign Language. To allow users to enter signs, a virtual keyboard has been developed. In addition, two prototype sign input devices are being researched, one based on video technology, and one based on datagloves. A prediction system has been developed, which, when combined with any of the input systems, reduces the effort required to enter signs, and also improves the recognition accuracy.
RESULTS--A prototype Sign PS system has been evaluated by users, with promising results. Very little instruction was required before the subjects were able to create signs and documents. In general the subjects were happy with the concept and the execution although some reservations were expressed. The glove and video systems are progressing well, with handshape recognition for both modules exceeding 90 percent for all 84 handshapes. Further parameters are currently being investigated. The prediction system has been shown to reduce user effort by as much as 96 percent.
FUTURE PLANS--Within the lifetime of the project, the document editor will be completed, and further research on the two input device prototypes will be done. The work could also be extended beyond the lifetime of the project with further work on the input devices, an improved font design, and a better display method, perhaps incorporating animation.
[162] ALADIN: ADVANCED LANGUAGE DEVICE FOR INTERACTION
Norman Alm, PhD; John L. Arnott, PhD; Iain Murray,
PhD
MicroCentre, Applied Computer Studies Division,
Department of Mathematics and Computer Science, University
of Dundee, Park Wynd, Dundee DD1 4HN Scotland, UK
Sponsor: Commission of the European Communities, Technology Iniative for Disabled and Elderly People, B-1049 Brussels, Belgium
PURPOSE--The ALADIN project is developing a novel, linguistically based software system which will enable a nonvocal, physically impaired person to hold effective conversations.
METHODOLOGY--The software will run on a wide range of commercially available hardware platforms (computer plus speech synthesiser). The system will include a model of conversational interaction which will provide the user with appropriate conversational material, prompts, and predicted utterances. An innovative interface is being developed which will help the user to navigate through a conversation with minimal attention to the interface and maximum attention on the other speaker. The user of this system will be persons who are non-vocal from birth through cerebral palsy, and also people who have permanently lost the ability to speak through degenerative conditions or accidents.
[163] THE FURTHER DEVELOPMENT OF TALKSBAC: A COMPUTER-BASED COMMUNICATION SYSTEM
Annalu Waller, PhD; Fiona Dennis, MCSLT; Nik J.
Whitehead, PhD; Alistair Y. Cairns, PhD; Jan K. Brodie,
MCSLT; Alan F. Newell, PhD; Ketron Morrison, MCSLT
Applied Computer Studies Division, University of Dundee,
Dundee, DD1 4HN, Scotland; Dundee Speech and Language
Therapy Service, Dundee DD1 9ND, Scotland, UK; email:
awaller@mic.dundee.ac.uk
Sponsor: The Leverhulme Trust, Tenovus Scotland
PURPOSE--The TalksBac project was awarded funding to continue a collaborative research project involving the University of Dundee and the Dundee Speech and Language Therapy Service.
The project is a continuation of the original TalksBac (Talking and Language Knowledge System for Better Aphasic Communication) project, which designed and evaluated a computer-based communication system, with four nonfluent Broca-type dysphasic adults. The project is staffed by a full-time researcher, a part-time researcher and a part-time speech and language therapist.
The broad aims of this project are to further develop the TalksBac system for dysphasic adults by implementing the improvements suggested in the outcomes of the previous project, to improve the training of carers in the choice of conversational material to be entered into the TalksBac database, and to evaluate the improvements in the new system by assessing its effectiveness in facilitating communication between client and partners.
METHODOLOGY--TalksBac is an augmentative communication system that uses predictive retrieval techniques to anticipate the sentences and narratives the dysphasic client may wish to use in conversation. Written in C++, TalksBac runs on a Macintosh PowerBook with an internal speech synthesizer. The first goal of this project was to further develop the user and caregiver software to encompass the findings of the previous project. The second goal was to develop a formal training procedure for caregivers to enable them to identify relevant conversational information which would be used by the dysphasic clients.
The second phase involves four dysphasic subjects using the system for 6 months. In order to measure the effectiveness of TalksBac, two types of evaluations are to be conducted at the end of this period. The first will focus on software: the speed of information retrieval and the performance of the prediction algorithm will be analyzed. The second evaluation will focus on the use of the software by the caregivers (i.e., the amount of data and the regularity in which the information has been entered into the system) and the users (the number of times the system is used in practice). The use of the system will also be evaluated in a similar way to the previous project in that the difference between unaided (without TalksBac) and aided (using TalksBac) communication will be analyzed using conversational profiles and videotaped conversations.
PROGRESS--The TalksBac software has been improved and tested and is now being used by four dysphasic adults. The caregivers are modifying and updating the databases on a regular basis. The dysphasic adults have been using their machines for a 6-month period, and the evaluation period is currently taking place.
RESULTS--Results of this project will be published in April 1997. Preliminary results are encouraging in terms of the increase in data entered into the databases. The implementation of the improved software has resulted in faster retrieval times.
[164] HUMAN FACTORS STUDIES IN EYE MOVEMENTS RELATED TO AAC HEAD MOVEMENT STUDIES
Richard Foulds, PhD; Arthur Joyce, PhD; Arup Khan,
BS
RERC on Augmentative and Alternative Communication,
Applied Science and Engineering Laboratories, University of
Delaware/A.I. duPont Institute, Wilmington, DE 19899
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 20202; Nemours Foundation, A.I. duPont Institute, Wilmington, DE 19899
PURPOSE--The use of eye movements as a method of interaction in augmentative communication has been explored for many years with limited success. Significant data exist on the ability of individuals with severe disabilities to coordinate their oculo-motor function with sufficient accuracy to use the line of gaze as an indicator of selection of a target. Instrumentation has been constructed using a camera to detect reflections of infrared light from the surfaces of the eye. This information allows the calculation of the line of gaze. Such a system can be used as a line-of-gaze typewriter or communication device.
The difficulties in the use of these instruments have been the human factors considerations associated with severe disability. Head movement is often unstable in individuals with disabilities. This project includes the development of an instrument which incorporates a pair of servo-controlled motorized mirrors that can follow the movement of the head in order to maintain the eye in the camera's field and make an eye tracking system that is considerably more appropriate for use with individuals with cerebral palsy and other disabling conditions.
METHODOLOGY--The major portions of this project involve the calibration and programming of the two servo-controlled mirrors connected to a 386-PC. A former graduate student wrote software libraries to control the mirrors, maintaining the eye in the camera's field of view. Rapid motion is still an unresolved problem: the eye-tracking software needs further development to compensate for the significantly more complex geometry.
FUTURE PLANS--Work is now underway to develop improved calibration techniques that will accommodate the movement of the head and allow for accurate calculation of the line of gaze.
[165] SINGLE SWITCH MOUSE CONTROL INTERFACE
Richard Mahoney, PhD; Andrew Moynahan
RERC on Rehabilitation Robotics, Applied Science and
Engineering Laboratories, University of Delaware/A.I. duPont
Institute, Wilmington, DE 19899; email:
mahoney@asel.udel.edu
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 20202; Nemours Foundation, A.I. duPont Institute, Wilmington, DE 19899
PURPOSE--The purpose of this project is to investigate the use of a intelligent agent within a single switch mouse control interface. It is believed that this type of interface will be more efficient than current single switch mouse emulation systems. With these, the burden of moving the mouse pointer is placed on the user, typically employing movement schemes based on a set of directions. If the user desires to move the mouse pointer to a window or icon on the screen, he/she is responsible for choosing the appropriate directions from the set of directions provided by the interface. Choosing a direction is typically achieved through scanning, which can be inefficient and forces the user to concentrate on the selection method rather than the desired task.
Our interface focuses on a goal-directed movement scheme. It operates on the assumption the user wants to move the mouse pointer to an object currently visible on the screen. The intelligent agent is responsible for detecting the potential targets on the screen and identifying the target desired by the user. Once identified, the interface is able to move the mouse pointer along a direct path to the destination point. This is possible because the interface is able to determine the mouse pointer's current location and the desired destination point.
METHODOLOGY--Input to the interface is achieved by means of a single switch. The user has the choice of depressing and holding down the switch or releasing the switch. Depressing and holding the switch sends a signal to the interface to begin moving the mouse pointer toward one of the targets on the screen. Releasing the switch stops all action. The user acts as a supervisor to the interface's actions. No action by the interface is allowed until the user depresses the switch. Once the user does so, the intelligent agent adjusts priority values associated with each target attempts to identify the target desired by the user. At this point the interface begins moving the mouse pointer along a direct path toward the target. The user is responsible for releasing the mouse pointer over the desired target.
PROGRESS--A preliminary planner has been developed and a single switch mouse control interface (SSMCI) has been implemented. An experiment has been carried out which tests the SSMCI against the task of target acquisition. A second interface, the directional arrows found on the numeric keypad, was also used in the experiment. A total of nine subjects, three of whom had disabilities, participated in the experiment.
RESULTS--Results include reduced average target capture times and reduced average number of switch activations when controlling the mouse pointer with the SSMCI as compared to the directional arrows.
FUTURE PLANS/IMPLICATIONS--Future plans include further refinement of the planner and testing the interface in a more complex environment. Also, we plan to develop an interface based on a single switch and an intelligent control interface for controlling a rehabilition robot. With the SSMCI, the intelligent agent was concerned with picking the desired target. While controlling a robot, the intelligent agent would be concerned with choosing the task desired by the user.
[166] DEVELOPMENT OF AAC SYSTEMS BASED ON PERSONAL COMPUTERS
Denise Peischl, BSE; John Dunaway, MS; Patrick Demasco,
MS; Alice Smith
RERC on Augmentative and Alternative Communication,
Applied Science and Engineering Laboratories, University of
Delaware/A.I. duPont Institute, Wilmington, DE 19899; email:
mineo@asel.udel.edu
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 20202; Nemours Foundation, A.I. duPont Institute, Wilmington, DE 19899
PURPOSE--This project focuses on the development of new techniques and strategies that advance the incorporation of technology in augmentative and alternative communication (AAC) based on personal and portable computers. In addition, results from other REC research efforts are incorporated into systems through this project. Finally, this project serves to develop systems for experimental evaluation of research techniques.
METHODOLOGY--The primary goal is to research and develop techniques to increase an individual's communication rate. This project focuses on enhancing a user's input process, particularly multimodal input, and increasing a user's vocabulary selection rate through a variety of prediction strategies.
PROGRESS--Multimodal Input in Computer Access This project focuses on issues related to the integration of two computer input strategies: speech input and head pointing. This project has worked in tandem with an evalutation project to provide a foundation for multimodal input integation for computer access and for future related research for multimodal input for AAC. A study that tests the hypothesis that each input technique is best suited to a set of related tasks is near completion. Head pointing is assummed to be advantageous for target acquistition tasks in a graphical interface, and speech recognition is presumed to be better for typing tasks involving text input. The development project produced a series of software programs used to conduct the study. Each program was designed to allow subjects to perform the same set of tasks using each of the input technologies. One program, MITE (Multimodal Input Text Editor), allowed subjects to transcribe paragraph sets, recorded a transaction record of the experiment, and collected performance data that was further analyzed by the evaluation project. A second program generated a series of target and destination objects in a graphical environment and allowed the user to select an object and drag it to a destination. This program also recorded a transaction of the experiment and collected performance data. In addition, the developement project worked with the evaluation project to design experiment protocols, develop the experiment environment, and conduct a series of pilot experiments.
ZapCom Work continues on another prototype communication system called ZapCom, the evolution of a photographic-quality, image-based system designed for individuals who have difficulty dealing with abstract symbol sets but work well with more life-like images. At this point, the system has been developed as a demonstration program.
A previous version, GraphCom, used gray-scale images, but ZapCom now uses color images captured by a small, inexpensive camera and put into the vocabulary set in a designated location. New images can be imported at any time. The vocabulary set is comprised of pages of images arranged in rows and columns. A VGA monitor displays one page of the vocabulary set at a time and the user constructs a message by selecting one or more images from the set.
ZapCom is a Windows application running on a laptop computer and integrates video capture and voice recording capabilities. The prototype version will be enclosed in a unit that contains the laptop computer, the video camera, battery supplies, and speakers. The video capture will be "point and shoot" procedure that allows a care provider to preview a capture window, take the picture and add it to the vocabulary dynamically.
FUTURE PLANS--Once development of ZapCom is complete, it will be transfered to the evaluation project for user testing. Additional multimodal research may continue to investigate the integration of multiple input strategies for augmentative communication. This project may also investigate communication applications for hand held, pen-based computers applications.
[167] EVALUATION OF HUMAN-SYSTEMS INTERACTION IN AAC
Beth Mineo, PhD; Denise Peischl, BSE, John Dunaway, MS,
Richard Lytton, MS
RERC on Augmentative and Alternative Communication,
Applied Science and Engineering Laboratories, University of
Delaware/A.I. duPont Institute, Wilmington, DE 19899; email:
mineo@asel.udel.edu
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 20202; Nemours Foundation, A.I. duPont Institute, Wilmington, DE 19899
PURPOSE--The challenge in technology development is the creation of innovative techniques that both maximize function and minimize the demands on the user. Realizing that each user of augmentative and alternative communication (AAC) brings a different set of skills to the communicative process, we must be careful about relying on unsubstantiated assumptions about the effectiveness of technology. Instead, we must begin to understand the efficacy of techniques in terms of the context in which they will be used and in terms of the capabilities that the user brings to the process. The goal of this project is to analyze the relationship between technological capability and functional use by individuals using AAC devices. Two projects are currently being evaluated: Language Facilitation through Graphics and Graphical Animation and Multimodal Input in Computer Access.
The results will provide guidance to those selecting and customizing AAC systems, as well as to manufacturers who are trying to make their products maximally responsive to the needs of people who rely on picture-based systems. The results will also provide feedback to research and developers with respect to the application of existing technology and to the design of strategies that incorporate cutting-edge technology.
PROGRESS--Language Facilitation through Graphics and Graphical Animation This project is investigating the representation of actions in two-dimensional forms. It is examining the relative efficacy of a number of approaches for representing movement, including static pictures, video, and animated pictures. The system presents a number of static graphics, animated graphics, and video clips to the subject. At present, evaluations are being performed with subjects (four years of age) from Easter Seals of DelMar. Preliminary data analysis has shown that across the different levels of graphical representations being used, digital video seems to be the most understandable. Another interesting observation is that for some subjects animation presented too much stimulus, making the task of identifying the requested "action" more difficult. This type of information will be valuable for clinicians, allowing them to assess individual skills prior to the recommendation of a particular augmentative communication system.
This evaluation study will continue to include 4-year olds but will also expand to include adults with cognitive disabilities.
Multimodal Input in Computer Access This project is investigating the issues surrounding the combination of two input technologies, speech recognition (using DragonDictate) and head pointing (using HeadMaster with WiVik). The hypothesis is that each device is better suited to one task (i.e., speech recognition for keyboard tasks and head pointing for mouse emulation). If this hypothesis is accurate, then one can infer that integration of the two input technologies will be beneficial.
Sixteen subjects without disabilities participated in the initial, three-phase study. In the first phase, subjects used each device for typing. In the second, subjects used each device separately to perform mouse-based pointing tasks. In the third phase subjects used both devices jointly for a spatial positioning task. Preliminary data concurred with our original hypothesis that speech recognition is faster for typing tasks then head pointing on an average of approximately 4 words per minute faster. Also, the average time for 25 repetitions of random spatial tasks showed that head pointing was faster than speech recognition. It took subjects, on average, approximately 2 min using head pointing and 7 min using speech recognition to complete these 25 repetitions. Interestingly, the time for conducting the repetitions using both modalities together was approximately 3 min.
The next phase of this evaluation study will be performed using subjects with disabilities. The protocol will be consistent with the one used with subjects without disabilities.
FUTURE PLANS--Following analysis of data collected from the studies, the results will be shared with the community of AAC professionals. In parallel, evaluation projects will be conducted on other projects developed within the ASEL.
In addition, efforts will be taken to explore the issues surrounding technology transfer, particularly for ZapCom and a previous project called Flexible Abbreviation Expansion. Potential manufacturers will be contacted to discuss the issues surrounding technology transfer.
[168] HUMAN FACTORS STUDIES IN EYE MOVEMENTS RELATED TO AAC HEAD MOUNTED UNIT
Richard Foulds, PhD; Arthur Joyce, PhD; Arup Khan,
BS
RERC on Augmentative and Alternative Communication,
Applied Science and Engineering Laboratories, University of
Delaware/A.I. duPont Institute, Wilmington, DE 19899
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 20202; Nemours Foundation, A.I. duPont Institute, Wilmington, DE 19899
PURPOSE--In order to overcome the many human factors obstacles in using eye movements to control AAC devices, a head-mounted unit is being developed that will compensate for many of the complications which affect the ultimate utility of an eye-tracking AAC system. A small, commercially available, head-mounted display (the Private Eye from Reflection Technologies) provides a convenient way to offer a computer display to an individual with a disability. The unit is worn on a head band and presents the image of the computer screen in the field of view of the wearer. This form of "heads up" display presents information that is independent of head movement.
METHODOLOGY--The primary goal of this project is to retrofit the display to incorporate a small camera capable of viewing the user's eye. The same software that is used to calculate the line of gaze in other eye tracking projects can be used to determine the line of gaze with respect to the "heads up" display. This system is potentially a portable eye gaze communication system that allows face-to-face communication.
PROGRESS--The optics and camera have been added to the display unit and the system tracks the corneal reflection and pupil center. Work is now underway to improve the optics and to teach the geometry of the system to the gaze calculation software so that it can translate the pupil center/corneal reflection vectors into gaze points.
FUTURE PLANS--The prototype system will be refined in the coming months. It must then be evaluated for its stability (the ability to tolerate the expected vibration and minor movements of the head mounted display.) Clinical studies using the entire system as a communication device will follow.
[169] THE APPLICATION OF NATURAL LANGUAGE PROCESSING TO AAC
Patrick Demasco, MS; Kathleen McCoy, PhD; Chris
Pennington, MS; John Gray, MS; Marjeta Cedilnik, BS; Terry
Carpenter, BS
RERC on Augmentative and Alternative Communication,
Applied Science and Engineering Laboratories, University of
Delaware/A.I. duPont Institute, Wilmington, DE 19899; email:
demasco@asel.udel.edu
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 20202; Nemours Foundation, A.I. duPont Institute, Wilmington, DE 19899
PURPOSE--The goal of this project is to investigate the application of natural language processing (NLP), a branch of artificial intelligence, to the development of more effective augmentative communication (AAC) systems.
METHODOLOGY--This work is based on the underlying concepts and model of communicative competence that describes augmentative communication system use on linguistic, operational, strategic, and social levels. Natural language processing provides the computational techniques necessary to give communication systems the capability to reason about lexical, syntactic, semantic, and pragmatic knowledge.
PROGRESS--Our previous work involved studies of language use in AAC systems and basic research into how natural language processing might be applied to AAC. One of the major results of our previous work was the development of the Compansion system which is a research prototype that takes uninflected content words (i.e., telegraphic input) and translates it into well-formed sentences. The prototype relies on both syntactic word-order information and semantic information about the meaning of individual words to determine an appropriate output sentence. We have begun to develop a more practical version of this system.
Additional work focuses on integrating pragmatic knowledge into a system based on conversational schemata. This powerful tool (named SchemaTalk) captures common language experiences and situations that can be accessed quickly to communicate effectively. The current focus involves determining methods for the system to help the user "schematize" his/her experiences so that appropriate language chunks can be accessed easily.
Also, we are developing an extensive object-oriented language database (Lexical Access Database or LAD) that will provide the detailed information necessary to support each natural language processing technique. The resulting database should provide information to allow efficient application of natural language processing techniques in the context of AAC.
FUTURE PLANS--Anticipated efforts include the integration of LAD into the compansion and schema-based projects, completion of the development and evaluation of a Compansion-based intelligent parser/generator (based on language data collected from AAC users), and significant improvement and testing of SchemaTalk, especially with respect to its user interface capabilities.
[170] SPATIALIZATION AND SPATIAL METAPHOR IN AAC
Patrick Demasco, MS; Blaise Liffick, PhD
RERC on Augmentative and Alternative Communication,
Applied Science and Engineering Laboratories, University of
Delaware/A.I. duPont Institute, Wilmington, DE 19899; email:
demasco@asel.udel.edu
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 20202; Nemours Foundation, A.I. duPont Institute, Wilmington, DE 19899
PURPOSE--In this project, we are interested in exploring new methods for the organization and access of language based on principles of spatialization and spatial metaphor that have emerged from the field of human-computer interaction.
METHODOLOGY--The predominant approaches to language organization that currently exist (e.g., levels) are largely based in the physical constraints of current AAC hardware such as screen size. Language is often organized to "fit" into the display or keyboard of the device. As a consequence, many users who have manual communication boards with hundreds of words are forced into the use of spelling, multiple levels, coding and/or predictive systems. As an alternative approach, we would like to consider technology that supports expanded information spaces as a means to provide more natural communication for individuals with severe communication impairments. The primary objectives of this project are to: 1) investigate the comparative usage of large manual word boards versus electronic systems by the same user; 2) develop a theoretical framework for describing new spatial metaphors in AAC; and 3) develop and evaluate prototypes of new systems that offer large language spaces that can be accessed in a multimodal fashion. We will utilize the technology available from the emerging field of virtual reality (VR) to create systems that consist of head-mounted displays providing the user with a view of the information space, a tracking system that will adjust the view based on head position, and a selection interface based on hand and/or eye pointing. In addition, the VR approach will be balanced with the design of systems that use conventional computer screens. The anticipated outcomes of this project are the contribution of a novel model of language organization and the demonstration, evaluation, and commercialization of systems that exploit this model.
PROGRESS--An application of the virtual word concept is under development. VAL (Virtual Access to the Lexicon), is intended to support spatial equivalence between manual and electronic systems. It also supports evolutionary changes in the word board structure, by allowing the user to access a large lexical database that stores words and their relationships to other words. Recent work has focused on collecting data on various mouse control methods for navigation of virtual keyboards and similar interfaces. A second prototype called VISOR (Visual Information Seeking of Reusable-Conversation) has also been developed. It provides a virtual environment for the storage and retrieval of narratives that are spatially indexed.
[171] SPEECH SYNTHESIS PROGRAM
Timothy H. Bunnell, PhD; Debra Yarrington, BA; Steve
Hoskins, BEE, MA; Keith Taylor, BS
RERC on Augmentative and Alternative Communication,
Applied Science and Engineering Laboratories, University of
Delaware/A.I. duPont Institute, Wilmington, DE 19899; email:
mineo@asel.udel.edu
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 20202; Nemours Foundation, A.I. duPont Institute, Wilmington, DE 19899
PURPOSE--The purpose of this program is to develop the software for the production of high quality, highly intelligible synthesized speech with an unlimited vocabulary. Areas of research in the program include the creation of a rule system for converting text into synthesized speech, the development of a method for changing the pitch and the duration of utterances, the development of a graphical user interface for the text-to-speech system, and the development of an automatic diphone extractor that allows for the customization of synthesized voices.
METHODOLOGY--The work in this program is based on diphone speech synthesis. Diphones are speech segments that run from the steady state of one phoneme through the transition between phonemes to the steady state of another phoneme. Diphones are obtained by recording someone saying words with the desired transitions in them, and then extracting and storing the transitions from the recorded words. Diphones can be appended together to create any word or phrase with the characteristics of the original speaker. Thus, diphones can be used to create a uniquely identifiable, natural sounding voice with an unlimited vocabulary.
PROGRESS--Previously the method for creating diphone libraries was completed. This process resulted in two synthesized voices: a male's and a child's. More recently, the focus of the program has been on improving the quality of the system that converts text to synthesized speech. The previous system calculated the phonemes, duration, and pitch on a word-by-word basis. The new system now performs a linguistic analysis on paragraphs of text and converts it into a number of linguistic representations: 1) the phoneme string with syllabification and stress information, 2) intonational features such as pitch accents and boundary tones, 3) prosodic phrasing, and 4) part-of-speech information.
The modeling of the intonation of questions, statements, exclamations, and the like, is greatly improved over the previous system and the modelling of the rhythmic characteristics is much more accurate. In addition, there is a great degree of user control over the prosodic characteristics of the speech. The previous system only allowed for control of the basic speaking rate and average pitch. The new system allows controls for pitch range, rate of declination, type of pitch accent, and many other options. The user may also override all of the automatic calculations of the synthesizer; this allows the addition of prosody not predictable from the text (such as emphasis on a particular word) or a simple means of correcting a wrong prosodic 'guess' of the text analysis.
A macro capability has been provided to allow the specification of complex controls with a minimal number of characters. Using this macro capability, preliminary commands to add emotion (happiness, sadness) to the prosodic characteristics of the speech have been developed. In order to facilitate the use of the text-to-speech system, a graphical user interface was developed under Windows 95. This greatly enhanced the program's ability to transfer technology. This program has also developed a method for automatically extracting diphones from recorded speech. The automatic extractor takes a set of recorded words and automatically determines the best instance of each diphone in those carrier words and then extracts and stores the diphone. The automatic extractor greatly reduces the amount of time and manpower needed to create a new set of diphones and thus a new synthesized voice.
RESULTS--The text-to-speech system is nearing completion. The rules for converting text to all the various linguistic representations have been completed and are being evaluated. The new system has been successfully ported to a Windows 95 environment within the graphical user interface of the old system. The user interface is being refined to allow easy control of the new system features. The automatic extractor was used on the same carrier words used in the manual development of the synthesized male voice, and the results were encouraging. In formal tests comparing the results of the automatic extractor and the manually extracted diphones, speech synthesized from automatically extracted diphones was consistently close in intelligibility and consistently rated as more natural sounding than the speech synthesized from the manually extracted diphones.
FUTURE PLANS--Future plans include thorough testing and refining of the text-to-speech system, thoroughly testing and refining the graphical user interface, testing and improving the efficiency of the automatic diphone extractor, and developing the methods for conveying emotions in the synthesized speech. Work will continue on the Windows 95 version of the synthesizer, to allow easy technology transfer. In addition to this application, DDE and DLL versions of the synthesizer will be developed and tested.
[172] EEG INTERFACE PROGRAM
Timothy H. Bunnell, PhD, James B. Polikoff, Winslow J.
Borkowski, Jr.
RERC on Augmentative and Alternative Communication,
Applied Science and Engineering Laboratories, University of
Delaware/A.I. duPont Institute, Wilmington, DE 19899; email:
mineo@asel.udel.edu
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 20202; Nemours Foundation, A.I. duPont Institute, Wilmington, DE 19899
PURPOSE--The object of this project is to explore the potential use of the P300 event related potential as a control signal in a computer interface for locked-in patients. There exists a significant population who, due to disease or injury, are totally paralyzed but have normal or near-normal brain function. In such cases, called Locked-in-Syndrome, the individual is aware of his or her surroundings, but has no way of communicating with the outside world. In cases where the person has even a slight degree of voluntary movement (e.g., eyebrow motion), it is possible to use that movement as a switch for controlling a computer. Likewise, when the person has good eye control, he or she can be fitted with an eye-tracking device to control cursor movement on a computer screen. In many cases, however, the individual may have no reliable voluntary motion to attach a switch to, and eye-movement may not be precise enough to use with an eye-tracking device. In such cases, the only possible method of communication would be to use electrical signals produced by the brain as a switching device for computer interaction. In order to achieve this, a reliable, detectable brain signal must be found.
Because of its robustness, we believe that an evoked electrical potential, called P300, may serve as a good candidate for an EEG-based computer interface. The P300 is a late positive wave that occurs between 250 and 800 ms after the onset of a meaningful stimulus. It was first reported in 1965 as a late positive component occurring in response to task-relevant stimuli. In this project, we will explore the possibility of using P300s to control cursor movement on a computer screen by presenting simultaneous visual target-detection tasks and measuring peak P300 amplitudes to targets occurring at each of four compass locations. Peak amplitudes are expected to be greatest for P300s in response to the direction that the subject is attending to.
METHODOLOGY--Subjects will be seated in a sound attenuated chamber facing a monochrome monitor 18 in (45.7 cm) distant. The central fixation point will be a cross. There will be four target arms (compass positions N, E, S, W) with a target (a cross) at the end of each arm and one cm from the central fixation point. Each stimulus will be presented for 250 ms with an interstimulus interval of either 750 or 1,000 ms. There are two different stimulus sets: in the first, each of the four target crosses will be replaced by an asterisk one at a time and in random order; in the second, a null-stimulus will be included, in which no asterisk appears. The subject will be instructed to fixate the central point and count the number of times one particular cross is replaced by an asterisk. The order of asterisk substitution will be random without replacement within each set of four (asterisk always appearing) or five (null included) stimuli. The target stimulus will occur with a probability of 0.25 in the first case, and 0.2 in the second. When a blink is detected (any signal beyond a preset threshold on the EOG channel), that stimulus trial will be discarded, and presented again later in the set. No set will be complete until at least one good (non-blink) trial is recorded for each target position. Thus, each set will consist of at least four or five trials (more if the subject blinks). Sessions will consist of 50 complete sets.
Data Acquisition. Grass silver-silver chloride electrodes will be placed according to the international 10-20 system at Fz, Cz, and Pz and referenced to bilateral (joined) earlobe electrodes. The EOG will be recorded from an electrode as SO2 (inferior and lateral to the right eye) also referenced to bilateral earlobes. The three EEG channels and single EOG channel will be amplified 50,000 times, bandpass filtered between 0.15 Hz and 150 Hz, and digitized (12-bit resolution) at a 300 Hz sampling rate on a 486 computer with an 8-channel DSP card. Data recording for each trial will begin 50 ms before presentation of the target stimulus and continue for a total of 650 ms. Thus, 600 ms of EEG data will be recorded for each channel following target onset. These data will be saved for subsequent analysis.
Data Analysis. Off-line analysis of collected data will model a real time process in which the computer estimates the direction in which the subject wishes to move the cursor, moves the cursor one step in the estimated direction, obtains another estimate of the desired direction, and so forth. It is in the nature of the task that each estimate must be independent of the last estimate since the subject must be free to change cursor direction at will. The estimated direction will be based on comparing P300 levels for targets on each arm of the cursor and selecting the largest P300 level as the most likely direction for cursor motion. This comparison can be made as soon as a single set (i.e., four target positions) has been obtained, or EEG activity for each target location can be summed over a series of sets to obtain a more stable P300 estimate.
RESULTS--Preliminary results showed that, overall, cursor movement based on P300 detection was correct about 50 percent of the time based on comparisons among peak levels within a single set of trials (chance is 25 percent). Overall average performance for our slowest subject was 0.13 bits per second (44.27 percent correct) and our fastest subject was 0.18 bits per second (51.92 percent correct). However, considering only the task in which target frequency was 0.2 for these two subjects, bit rates were 0.15 and 0.27 respectively. Accuracy increased when successive sets of trials were summed before comparing P300 levels, but this gain in accuracy was accompanied by an increase in the amount of time to make a decision. In no data that we have examined to date have we observed an instance in which the trade-off between accuracy of P300 detection and time would result in advantages for summing trials; the cost in time far exceeds the benefits of accuracy in this task.
FUTURE PLANS--Averaging over trials in the present task does not appear to be a productive way to improve bit rate. However, varying other task variables like target frequency and presentation rate may lead to moderate improvements in the accuracy with which P300 events are detected. In future studies we will continue to explore these and other task variables to find conditions which lead to optimum performance. As a signal for control of communication devices and interfaces, the P300 has several limitations. The most serious of these is the relatively low bit rate associated with its use. However, for some potential users, this low bit rate may still exceed the rates available via other communication channels, and at present, communication rates associated with P300 detection seem equivalent to those associated with the detection of other brain events or states.
[173] RESEARCH IN INTERFACE METHODOLOGIES FOR AAC
Richard Foulds, PhD; Roman Erenshteyn, PhD; Garland
Stern, PhD; Lynn Messing, PhD; Arthur Joyce, PhD; Scott
Galuska, MS; Pavel Laskov, MS
RERC on Augmentative and Alternative Communication,
Applied Science and Engineering Laboratories, University of
Delaware/A.I. duPont Institute, Wilmington, DE 19899; email:
foulds@asel.udel.edu
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 20202; Nemours Foundation, A.I. duPont Institute, Wilmington, DE 19899
PURPOSE--The goal of this research is to explore new and innovative methods for human interaction with AAC devices. This program is comprised of projects in two areas: basic research in developing methods for human gesture to be used as an interface for AAC devices and a human factors study in the use of eye movements to control an AAC device.
METHODOLOGY--Gesture Project: This project explores the use of gloves and other sensors (such as those used in virtual reality applications) as input devices to control an AAC device. Input taken from the gloves and sensors are fed into trained neural networks which attempt to extract meaning based on hand shape, hand position, and movement in space. Both formal gestural systems, such as American Sign Language (ASL) and informal gestural systems are being studied.
Eye Movement Project: This project re-examines the use of eye gaze to control an AAC device. The first part of the project-attempts to define human factors parameters related to eye gaze. Latter parts of the project will examine the trade-offs in using head-mounted versus remote trackers in terms of accuracy and usability. The final phase of the project involves studying the effects of oculomotor disabilities on the performance of individuals with disabilities in their use of an AAC device using eye gaze.
PROGRESS--Gesture Project: The current work in ASL recognition is focused on improving the accuracy of the system, and its adjustment to handling signing at natural speed. Two techniques have been developed which provide an accuracy of 85-90 percent and significantly reduce the training time of the system. The motion recognition unit has been designed and tested on 8 basic motions used in ASL and yielded satisfactory performance.
Additionally, work has proceeded on synthesis of Signed English in the form of an animated virtual signer. Individual signs are described tersely in terms of handshape, position, and attitude of the hands, and motion of the hands during the sign. It is expected that the database for sign synthesis will be used as a part of the full recognition system. This will provide a straightforward way to extend the recognition repertoire.
Eye Movement Project: Progress has been made in the use of motorized mirrors to track eye movement, allowing a limited range of head motion in the user. Work on developing a head-mounted tracker is also being pursued.
[174] AUGMENTATIVE AND ALTERNATIVE COMMUNICATION TECHNICAL ASSISTANCE AND OUTREACH PROGRAM
Julia Mercier, BA; Ron Sibert, BS; Beth Mineo, PhD;
Richard Foulds, PhD
RERC on Augmentative and Alternative Communication,
Applied Science and Engineering Laboratories, University of
Delaware/A.I duPont Institute, Wilmington, DE 19899; email:
foulds@asel.udel.edu;
mercier@asel.udel.edu
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 20202; Nemours Foundation, A.I. duPont Institute, Wilmington, DE 19899
PURPOSE--Our goal is to serve as a leading source of materials and assistance to professionals, consumers, families, and agencies dealing with augmentative communication (AAC) and to disseminate technical reports and articles on research results obtained by ASEL staff. The program participates in activities to promote dissemination in the areas of technical assistance, consumer advocacy efforts, support to manufacturers, and support to other researchers.
METHODOLOGY--Dissemination tools, including publications, a World Wide Web site, booklets, and special reports are developed and disseminated. The Information Program also offers support to "Tech Act" state projects in the AAC arena.
PROGRESS--The researchers in AAC have produced 15 publications for dissemination through the Information Program. Inquiries from other researchers, clinicians, therapists, consumers, teachers, and families are answered with a personal response and appropriate information materials. The 1996 Guide to Augmentative & Alternative Communication Devices has been published, along with accompanying vendor information notebook and color slide set, and distributed to over 500 interested parties. A World Wide Web site, "AT On-Line," has been established as an electronic information base. ASEL was also selected to host the International Conference on Spoken Language Processing in October 1996.
[175] ENGAGING, RECRUITING, AND RETAINING STUDENTS WITH DISABILITIES IN SCIENCE, ENGINEERING, AND MATH
Kenneth Barner, PhD; Richard Foulds, PhD; Thomas Hammer,
MEd; Anna Phalangas, BS; Richard Schumeyer, MS; Thomas Way,
BS; Jason Fritz, BS; Judy Trefsger
SEM Tactile and Haptic Interface Project, Applied
Science and Engineering Laboratories, University of
Delaware/A.I. duPont Institute, Wilmington, DE 19899; email:
sem-info@asel.udel.edu;
barner@udel.edu
Sponsor: National Science Foundation, Arlington, VA 22230; Nemours Foundation, A.I. duPont Institute, Wilmington, DE 19899
PURPOSE--Individuals with disabilities are currently underrepresented in science, engineering, and math (SEM) academic programs and professions. While this underrepresentation is due to many factors, several impediments are clear: attitudinal, physical, and curriculum barriers combine to cause individuals with disabilities to be stymied in both SEM education and professions. Currently, attitudinal barriers reside not only in school counselors, teachers, and employers, but also in the students themselves and their family members. Similarly, physical barriers manifest themselves not only in the way classrooms are set up, but also in the way information is conveyed in both lectures and experiments. This project specifically targets each of these attitudinal, physical, and curriculum barriers, and is designed to allow individuals with disabilities to flourish in SEM.
METHODOLOGY--Individual attitudinal barriers are broken down by providing positive SEM experiences and through a mentoring program. The mentoring program utilizes the Internet as a "distance free" pathway for communications, over which students and mentors can communicate. The attitudes of school counselors and teachers, which currently discourage students with disabilities from pursuing SEM curricula, are changed through education and abilities demonstrations. Similar methods are used to educate and change the attitudes of family members and employers. Physical and informational barriers are broken down through the design and development of new information access methods and virtual laboratories, which are both physical and information barrier free. New information access methods include tactile and haptic interfaces for individuals with visual impairments.
PROGRESS--The Internet mentoring program enrollment currently stands at 25 students. All the students have either physical, sensory, or learning disabilities. A matching number of mentors, approximately half of whom have disabilities, are participating in the program and paired one-to-one with students. In addition to the online mentoring activities, students participated in a Summer Science Fest program that featured team building, computer training, aerospace activities, biology/nature education, and information about college, including admissions, ADA services, academic information, and campus tours. To address the problems affecting teachers, counselors, and families, the project sponsored several workshops and conferences focusing on learning disabilities, assistive technology, education, and college transition. Prototype tactile and haptic visualization systems have also been developed. The tactile system utilizes image manipulation to produce comprehendible tactile output of images on medium such as capsule paper while the haptic system is based on the PHANToM force reflecting device. Interested parties are encouraged to send e-mail to sem-info@asel.udel.edu or examine the project web page (http://www.ee.udel.edu/InfoAccess/) for more information.
[176] SPEECH PROCESSING PROGRAM
Timothy H. Bunnell, PhD; James B. Polikoff, MS; Shirley
M. Peters, MS; Xavier Menendez, PhD
RERC on Augmentative and Alternative Communication,
Applied Science and Engineering Laboratories, University of
Delaware/A.I. duPont Institute, Wilmington, DE 19899
Sponsor: National Institute on Disability and Rehabilitation Research, U.S. Department of Education, Washington, DC 20202; Nemours Foundation, A.I. duPont Institute, Wilmington, DE 19899
PURPOSE--The purpose of this project is to integrate computer hardware and software that will implement speech enhancement algorithms developed for dysarthric speech as a real-time prototype speech prosthesis. The device will accept speech produced by a person with a speech disorder, process the speech to make it more natural sounding, and then replay the processed speech on command. The prototype under development is to serve as a test bed for implementing new speech processing algorithms and for studying how users will interact with such a device.
METHODOLOGY--This program is based on ongoing work in the Speech Processing Laboratory to develop signal processing techniques capable of improving the intelligibility and naturalness of disordered speech. These techniques involve adjusting the timing of the speech as well as adjusting its spectral properties. Timing adjustments are performed by simply cutting out unwanted portions of the speech, or lengthening (by repeating) sections of speech that are too short. This form of signal processing is computationally simple and fast and therefore attractive from a practical standpoint. More complex rule-based systems are also being developed that will require recognition of general acoustic speech patterns to determine how segments should be shortened or lengthened. These systems will be able to automatically decide which parts of the original speech are important to keep and which can be safely discarded without losing important information.
Once optimal timing has been accomplished, spectral characteristics of the speech can be adjusted to further enhance intelligibility. We are currently working on an implementation of an interface for the speech prosthesis between a PC and DSP card. Perceptual experiments have been performed to test the intelligibility of the disordered speech both before and after speech processing. In a typical experiment, normal-hearing subjects listened to samples of original and processed speech in a sound-attenuated chamber. In tests of segmental intelligibility, subjects listened to short nonsense sentences and chose the words they thought they heard from a closed response set differing on a single phoneme. These experiments helped to identify which speech production errors are common for specific disordered talkers, as well as which type of articulations were helped (or hindered) by speech processing.
RESULTS--From the first set of studies we concluded that time-adjustment leads to significantly better sounding speech and small but significant improvements in intelligibility for some phonemes. However, improvements in intelligibility were sometimes offset by artifacts of the signal processing techniques used for timing adjustment. A new version of the software for timing adjustment has been developed which minimizes such processing artifacts and at the same time incorporates a simple heuristic for determining which segments to alter in adjusting the timing of the signal.
[177] EFFECTIVENESS OF USING VOICE RECOGNITION SYSTEMS
Ava-Lee Kotler, MS; Cynthia Tam, BSc; Pat Stoddart,
BA
Bloorview MacMillan Centre, Toronto, ON M4G 1R8,
Canada
Sponsor: Ontario Ministry of Health, Neurodevelopmental Clinical Research Unit, Hamilton, ON L8S 4K1
PURPOSE--The purpose of this pilot study is to evaluate the use of voice recognition to create text and the effects of discrete utterance speech on participants' voices. Components that will be evaluated include speed of creating text, fatigue, ability of the system to complete tasks, and ease of use. In addition, vocal characteristics will be analyzed to determine whether the voice has been affected from use of voice recognition.
METHODOLOGY--Participants will be clients who have been prescribed voice recognition systems to use as writing aids. Six will be recruited for this preliminary study and will be asked to complete various typing tasks and to answer questions regarding the use of their system. Speed and accuracy of text generation, and the voice recognition system's accuracy, will be calculated. Tape recordings of speech samples will be obtained in order to evaluate voice quality. Responses to interview questions will be analyzed descriptively.
PROGRESS--Four participants have completed the above protocol; each has a different diagnosis (e.g., muscular dystrophy, upper motor neuron disease, scleroderma, and thoracic output syndrome).
PRELIMINARY RESULTS--Data has only recently been obtained and analysis is in process.
FUTURE PLANS--A future research study with a larger sample will be conducted.
[178] ESTABLISHMENT OF A DATABASE FOR IDENTIFICATION OF AUGMENTATIVE COMMUNICATION AID USERS AND FACILITATORS WILLING TO PARTICIPATE IN RESEARCH
Bernard M. O'Keefe, PhD; Jeff Jutai, PhD; Reinhard
Schuller, MSc; Paul Marshall; Peter Lindsay, PhD
Department of Speech-Language Pathology, University of
Toronto, Toronto, ON L6L 5N8; Bloorview MacMillan Centre,
Toronto, ON M4G 1R8; Ontario Institute for Studies in
Education, Toronto, ON M5S 1V6; email: b.okeefe@utoronto.ca;
ortcjj@oise.on.ca;
plindsay@oise.on.ca
Sponsor: Ontario Rehabilitation Technology Consortium, Toronto, ON M4G 1R8; Industry Canada, Ottawa, ON K2H 8S2
PURPOSE--The purpose of this pilot project is to establish a database containing the names of individuals, age 16 or over, with cerebral palsy who do not speak or are unable to speak successfully in at least some situations important to them. The database is being established so that researchers interested in improving the communication systems of such individuals will be able to call upon those persons to: 1) help determine what research needs to be done; 2) participate in the design and carrying out of projects; and/or 3) cooperate in interpreting the data gathered during projects. The database will also help ensure that consumers are represented at all levels of research and provide employment for persons with disabilities. It is expected that the results of this project will provide a model for extension to other disability groups.
METHODOLOGY--A draft questionnaire was designed from existing literature. Key informant interviews were held with consumers and their facilitators who helped modify the draft. The modified draft was examined by representatives of consumer groups, a group of scientific researchers, and manufacturers of augmentative and alternative communication (AAC) devices, and further changes were made. The revised questionnaire was pretested with AAC users and final revisions were made. The database was then designed, using MicroSoft Access and WiViK access software, by a design team which included paid employees of UserNet, a company whose employees are persons with disabilities. Cooperating agencies provided names of potential participants who were contacted and provided questionnaires.
PROGRESS--Early questionnaire returns have been encouraging. Additional major consumer groups are providing contacts. Data on returned questionnaires are successfully being entered into the database primarily by UserNet employees.
FUTURE PLANS--The marketing and business plans will be completed by Autumn 1996, when the database is anticipated to begin operations. Reviews after 1 and 2 years of operation will be conducted in order to determine whether the database is succeeding and should be expanded to other disability groups. When the database is in operation, a review panel will evaluate requests for specific lists of names from scientific researchers, market researchers, and consumer group representatives based on a published set of criteria. If the project is approved, database employees will contact the individuals on the list and ask them if they are willing to be contacted about the approved project; only when the individual agrees may the name of the individual be released to the requesting source.
[179] HOME AUTOMATION AND WORKPLACE INTEGRATION
Andre Wisaksana, MS, MEng; Geb Verburg, MA; Anastasia
Cheetham, BASc; Morris Milner, PhD, PEng
Institute of Biomedical Engineering, University of
Toronto; Rehabilitation Engineering Department, Bloorview
MacMillan Centre (MacMillan site), Toronto, Ontario; email:
verburg@ecf.utoronto.ca;
ortcgv@oise.on.ca
Sponsor: Ontario Rehabilitation Technology Consortium funded by the Ontario Ministry of Health; the Natural Sciences and Engineering Council; the Hospital for Sick Children Foundation; The National Strategy for the Integration of Persons and Disabilities; Industry Canada
PURPOSE--Assistive technologies have significant benefits for persons with disabilities. Presently, these benefits are offset by the effort required to integrate these independent and often proprietary devices into coherent systems that address the needs of their user. The main obstacle to integration is the lack of interoperability due to devices not speaking the same "language." Here, the technical analogy of a language is a communications standard, under which devices that comply to the same standard are able to communicate with each other.
PROGRESS--In Europe, the Multiple-Master Multiple-Slave (M3S) standard has been developed and proposed as the communications standard for assistive technologies. It is presently being considered as a preliminary standard by the International Standards Organization (ISO).
A wide variety of other nonrehabilitation specific devices/systems also exists that may assist persons with disabilities, such as personal computers and environmental controls. In order to incorporate the full functionality, and hence benefits, of these other technologies, a gateway is needed. Gateways are functional bridges between two systems, allowing the user to interact with both.
This project also cultivated closer ties with TPD-TNO, the main European M3S developer, and other European organizations. This has already resulted in the definition of future collaborative projects, such as TIDE (Telematics Initiative for Disabled and Elderly Persons), INCONTROL (Innovative Control of Mobility, Manipulation, and the Environment), and ARTISTE (Access to Rehabilitation Technology Information Services and Transfer of Expertise).
FUTURE PLANS--Continued development is needed to further refine the designs and minimize their size. Efforts are also underway to set up clinical and client demonstration sites. This will provide much needed user input and feedback. In addition, further evaluation is needed regarding system safety and reliability, and a rational and objective method of evaluating M3S compliance needs to be defined.
[180] HAMLET--SIMULATING EMOTION IN SYNTHETIC SPEECH
Iain Murray, PhD; John Arnott, PhD
MicroCentre, Applied Computer Studies Division,
Department of Mathematics and Computer Science, University
of Dundee, Park Wynd, Dundee DD1 4HN Scotland, UK; email:
imurray@mic.dundee.ac.uk
Sponsor: None at present
PURPOSE--This project aims to derive a set of rules to include vocal emotion effects in synthetic speech produced by rule.
METHODOLOGY--A basic set of rules were derived from the existing literature on human vocal emotion to produce a prototype system (HAMLET). Later analysis of actor recordings were used to enhance and extend the system.
PROGRESS--A Windows version has been developed.
[181] DIRECT BRAIN INTERFACE BASED ON DETECTION OF EVENT-RELATED POTENTIALS
Simon P. Levine, PhD; Jane E. Huggins, MS; Spencer L.
BeMent, PhD; Lori A. Schuh, MD; Ramesh K. Kushwaha, PhD;
Donald A. Ross, MD; Mitchell M. Rohde, MS
Physical Medicine and Rehabilitation, Biomedical
Engineering, Electrical Engineering and Computer Science,
Neurology, Neurosurgery, The University of Michigan, Ann
Arbor, Michigan 48109-0032; email: silevine@umich.edu;
janeh@umich.edu
spence@umich.edu;
ramesh@umich.edu;
daross@umich.edu;
bovine@umich.edu
Sponsor: None listed
PURPOSE--This research explores the detection and use of intracranial subdural event-related potentials (ERPs) to demonstrate the feasibility of a direct brain interface for people with disabilities. Many people with disabilities could potentially benefit from an interface that does not require physical movement and accepts commands directly from the brain.
The short-term goal of this research is the development of methods capable of accurately detecting the ERP related to a simple voluntary action. Such methods would then be used as the basis of a direct brain interface which would detect the ERP corresponding to the planning of a voluntary action and produce a single switch closure in response.
The long-term goal of this research is the development of a direct brain interface which could be used by people with severe disabilities (such as locked-in syndrome) to operate complex external devices without the need for physical movement.
METHODOLOGY--The subjects in this study are patients in an epilepsy surgery program who have had subdural electrodes placed on their cortical surface for presurgical monitoring. Electrocorticograms (ECoG) are recorded from these electrodes while the subjects repeat simple voluntary actions at intervals of 3 to 10 s. The time at which each repetition of the action was performed is recorded, using either a simple switch, microphone, or electromography (EMG) electrode. For each subject/action data set, triggered averaging is used to produce an average ECoG segment corresponding to the repetition of the action for each electrode location. These averages are then evaluated to determine which electrode locations recorded ERP's corresponding to the voluntary action.
Currently, cross-correlation is being evaluated for the detection of the ERP's. First, an ERP template is cross-correlated with unprocessed ECoG from the same electrode location. Then an experimentally determined threshold is applied to the cross-correlation statistic and points which exceed this threshold are considered possible detection points. The performance of cross-correlation as a detection method is evaluated by comparison of the detection points with the actual times at which repetitions occurred as documented by the trigger channel.
PROGRESS--Data have been collected from seven subjects. Identification of ERPs and evaluation of a cross-correlation-based detection method was performed with ECoG from these subjects. The software was updated to permit selection of the template duration based on signal content. Evaluation of cross-correlation-based detection using templates whose duration is selected as a function of the signal content of the template is in initial stages.
RESULTS--ECoGs from 28 subject/action data sets have been analyzed. Cross-correlation-detection methods were evaluated, using averages which began 2.5 s prior to the trigger point and ended 1.5 s after it. For three subject/action data sets, all from different subjects, one data set was found for which the percentage of actions correctly detected by the cross-correlation-detection method (the hit percentage) was greater than 95 percent, and the percentage of incorrect detections (the false positive percentage) was less than 5.
FUTURE PLANS--Evaluation of the cross-correlation-detection methods will continue with the comparison of fixed-length templates and templates whose length is determined based on signal content. Analysis of the effects of low-pass filtering of the ECoG prior to cross-correlation is also planned. Preparations are being made for intraoperative data collection during which electrodes can be placed over the motor cortex. Intraoperative data collection would increase experimental control with regard to electrode location, which are currently placed solely for the purpose of presurgical monitoring.
[182] RELATIONSHIPS AMONG AGE AT ONSET, ADEQUACY OF PERSONAL ASSISTANCE, NEGATIVE HEALTH INCIDENTS, AND HEALTH CARE UTILIZATION FOR PERSONS WITH PHYSICAL DISABILITIES
Margaret A. Nosek, PhD; Carol A. Howland, MPH; C. Don
Rossi, MS; Kathy L. Meroney, BA
The Center for Research on Women with Disabilities,
The Department of Physical Medicine and Rehabilitation, and
The Institute for Rehabilitation and Research, Baylor
College of Medicine, Houston, TX 77046
Sponsor: National Institutes of Health, National Center for Medical Rehabilitation Research, Bethesda, MD 20892
PURPOSE--The purpose of this longitudinal study is to determine the strength of relationships among use of personal assistance services for activities of daily living, health status, and use of health care services by persons with a variety of severe physical disabilities.
METHODOLOGY--Survey packets were mailed to 120 subjects with the objective of having at least 100 complete the study, allowing for attrition. Data collected was used to construct a profile on each participant and his/her health service use and health conditions over a 12-month period. These profiles were used to identify differences between participants based upon whether personal assistance services are provided exclusively by family members or by nonproviders alone or supplementing family assistance.
In year 2, open-ended, qualitative interviews were conducted with 20 participants who represented two subsets of the original sample: 5 whose scores on the Personal Assistance Satisfaction Index (PASI) fall into the top quartile and 5 in the bottom quartile, and (2) 5 whose total number of negative health incidents fall into the top and 5 in the bottom quartile. The sample for this segment of the study was strictly limited to 20 participants to allow in-depth exploration of PAS and health issues.
PROGRESS--One hundred persons, aged 18 to 65, living independently in the community, who use at least 1 hour of personal assistance daily, completed weekly checklists for 1 year, recording any changes in their personal assistance or health status, any visits to hospitals, emergency rooms, or physicians, and, when health incidents occurred, any effects on productivity and levels of distress. Data analysis has been completed on the initial PAS and Health Study questionnaire, as well as the monthly health and personal assistance incident checklists. Hypothesis testing is ongoing.
RESULTS--Participants were generally satisfied with the quality of their personal assistance services, possibly because most participants had many years of experience to find an arrangement that best met their needs. However, they were less satisfied with the availability of attendants and costs of services. Overall, participants were in very good health, with few or no acute illnesses in a year. More than one third reported no acute health incidents, and another third had 20 or fewer days of poor health. In year 3, the total number of health and personal assistance incidents that occurred over the 12-month reporting period was analyzed and showed that the top five personal assistance incidents included: assistant did not perform a task (299), assistant did not perform a task in the manner requested (263), assistant did not show up/patient had to find a back-up (186), assistant arrived late for work (141), and assistant was verbally abusive: yelling, cursing, or threatening (131). The top five health incidents that occurred were: pressure sores (1147), urinary tract infections (832), joint contractures (217), fractures (189), and colds or flu (115). Same day occurrences of health and personal assistance incidents were counted in order to provide a frequency of paired incidents. The kinds of paired incidents were noted.
Overall, quality of personal assistance had an impact on the health status of persons with severe physical disabilities in maintaining mobility and nutritional status. There was a correlation between the mean number of monthly health and personal assistance incidents. However, the majority of health incidents were not associated with personal assistance incidents. Only 4.5 percent of health incidents occurred within a week of a personal assistance incident. A small subset of participants had a large occurrence of health incidents, personal assistance incidents, and health incidents that could be attributed to inadequate personal assistance services. Characteristics of this subset need to be explored further.
[183] INCREASING THE CAPACITY OF INDEPENDENT LIVING CENTERS TO SERVE MINORITY POPULATIONS
Margaret A. Nosek, PhD; Carol A. Howland, MPH
The Center for Research on Women with Disabilities,
The Department of Physical Medicine and Rehabilitation,
Baylor College of Medicine, Houston, TX 77046
Sponsor: National Institutes of Health, National Center for Medical Rehabilitation Research, Bethesda, MD 20892
PURPOSE--The Center for Research on Women with Disabilities is collaborating with the ILRU Research and Training Center on Independent Living to conduct a study to find out how independent living centers could improve the effectiveness of their services for minorities with disabilities.
METHODOLOGY--We are focusing on a different minority group for each year of this 5-year study. This year we will be looking at issues facing African-Americans. We will begin each year with focus groups of consumers to help us identify the factors that affect their independence.
PROGRESS--Recruitment of African-American men and women with a variety of physical disabilities has begun. Focus groups are being conducted to determine how independent living centers can better serve their needs.
FUTURE PLANS--The information gathered from the focus groups will be presented to independent living centers around the country for comment and analysis of how their services address these needs and barriers they encounter in delivering those services to this population.
[184] THE ACCESSIBILITY OF PRIMARY CARE PHYSICIANS' OFFICES FOR PEOPLE WITH DISABILITIES: AN ANALYSIS OF COMPLIANCE WITH THE AMERICANS WITH DISABILITIES ACT
Ellen W. Grabois JD, LLM; C. Don Rossi, MS; Margaret A.
Nosek, PhD
The Center for Research on Women with Disabilities,
The Department of Physical Medicine and Rehabilitation,
Baylor College of Medicine, Houston, TX 77046
Sponsor: National Institutes of Health, National Center for Medical Rehabilitation Research, Bethesda, MD 20892
PURPOSE--The purpose of this project is to answer two questions: 1) What is the level of accessibility of private physician offices to patients with physical disabilities? and 2) Are physicians in compliance with the Americans with Disabilities Act of 1990 (ADA)? Physician offices are public accommodations and must comply with Title III of the ADA.
METHODOLOGY--A random sample of 220 primary care physicians in general practice, family practice, internal medicine, and obstetrics-gynecology were chosen from the Harris County (TX) Medical Society roster. Each was sent a 57-item questionnaire with 136 variables related to accessibility, the ADA, and care and treatment of patients with physical disabilities.
PROGRESS--Data analysis for the project has been completed. Sixty-two physicians, or 28.2 percent, responded to the questionnaire. Frequencies were recorded for each question.
RESULTS--In general, physicians reported a high level of compliance with 17 readily achievable structural features in their offices. The majority (82.3 percent) reported no difficulty in serving patients with disabilities. However, 11.3 percent reported they were unable to serve 1-2 disabled patients in the last 12 months. Also, 19.4 percent said that patients with disabilities were too difficult to treat or handle, and 22.4 percent felt more comfortable referring patients with a disability to another physician. In addition, 3.2 percent of patients were unable to enter premises because of physical barriers. These results indicate a strong need for continuing medical education on the requirements of the ADA, and strategies for assisting physicians to come into compliance with it.
[185] A CURRICULUM FOR TRAINING PHYSICIANS IN REPRODUCTIVE HEALTH CARE FOR WOMEN WITH PHYSICAL DISABILITIES
Margaret A. Nosek, PhD; Carol A. Howland, MPH; Catherine
Clubb Foley, PhD; Ellen W. Grabois JD, LLM
The Center for Research on Women with Disabilities, The
Department of Physical Medicine and Rehabilitation, and The
Institute for Rehabilitation and Research, Baylor College of
Medicine, Houston, TX 77046
Sponsor: National Institutes of Health, National Center for Medical Rehabilitation Research, Bethesda, MD 20892
PURPOSE--The purpose of this project is to enhance the ability of primary care physicians to meet the reproductive health care needs of women with disabilities. Such women face serious obstacles to receiving such services. Our Center is in the final phase of a 3-year national study to examine psychosocial behaviors of women with physical disabilities. During the course of this study, several problems with reproductive health surfaced with unexpected strength. These problems include 1) lack of physical access to physicians' offices and equipment; 2) gaps in the knowledge of physicians and other health professionals about how disabilities affect reproductive health care needs; 3) assumptions by health care professionals that such women do not need reproductive health care; 4) deficits in knowledge and faulty beliefs of these women about the functioning of their bodies and their need for reproductive health care services; and 5) problems inherent in health care service systems. In our national survey, we found that these women face certain unique problems in sexual response, pregnancy, delivery, and the detection of sexually transmitted diseases, breast cancer, and cervical cancer.
METHODOLOGY--A case-study-based medical education curriculum for primary care physicians was developed that offered the most current information on serving the reproductive health care needs of women with disabilities. In developing this curriculum, our Center used findings from the national study, information gathered from 15 faculty members at Baylor College of Medicine, recommendations from a national medical advisory committee, and the expertise of the staff for this project.
PROGRESS--A training curriculum has been developed by our Center. This training was offered to the faculty and residents in the Baylor College of Medicine departments of Family Practice, Community Medicine, and Obstetrics/Gynecology. Four completed trainings have been conducted, and four more are scheduled. At the end of the project year, a continuing medical education seminar will be offered that will be advertised to primary care physicians throughout the Houston area and nationally.
FUTURE PLANS--National dissemination of training materials to primary care physicians is planned.
[186] HEALTH PROMOTION FOR WOMEN WITH PHYSICAL DISABILITIES
Margaret A. Nosek, PhD; Mary Ellen Young, PhD; Carol A.
Howland, MPH; Gail Chanpong, MS
The Center for Research on Women with Disabilities, The
Department of Physical Medicine and Rehabilitation, and The
Institute for Rehabilitation and Research, Baylor College of
Medicine, Houston, TX 77046
Sponsor: National Institutes of Health, National Center for Medical Rehabilitation Research, Bethesda, MD 20892
PURPOSE--The concept of wellness in the context of physical disability among women has only recently been introduced into the field of health promotion. The purpose of this project is to develop an intervention to enhance wellness among such women based on expanded theoretical models and measures of health-promoting behaviors that accommodate some of their unique life circumstances. Findings from our recently completed national survey indicate that segments of the population of women with physical disabilities are at a higher risk for certain acute and chronic conditions, limited access to preventive health services, negative social attitudes toward their potential for fitness and wellness, and reduced motivational factors affecting health promoting behaviors, such as self-esteem, self-efficacy, and body image. There is a pressing need for the development of health promotion programming that is responsive to the needs of women with significant functional limitations and an accompanying research protocol to measure the effectiveness of such programming.
The specific aims of this study are to
This research will be grounded in the Transtheoretical Model which identifies five stages that characterize an individual's willingness to change a given behavior: precontemplation, contemplation, preparation, action, and maintenance.
METHODOLOGY--To accomplish the aims of this study, we will begin with qualitative focus groups and individual interviews to understand how women with physical disabilities define wellness and practice health promoting behaviors.
PROGRESS--Recruitment of participants has begun and focus groups will be conducted.
FUTURE PLANS--In the next phase, we will use quantitative methodology to establish a baseline for the perceived health status, health promoting behaviors, and stage of change of a sample of 400 women with a range of physical disabilities and severity levels. We will then develop a motivational intervention, consisting of a newsletter, a wellness hotline, a fax-on-demand service, and a home page on the World Wide Web, to offer information and encourage action for women who may be in early stages of change or believe that their disability prevents them from engaging in health promoting activities. In the final phase, we will develop and test the feasibility of a wellness action group program that will focus on individualized goal-setting, peer support, and enhanced self-efficacy for pursuing healthier living.
[187] A STUDY OF POLICY BARRIERS IMPEDING USE OF ASSISTIVE TECHNOLOGY BY PERSONS AGING WITH DISABILITIES
Phoebe S. Liebig, PhD; Debra J. Sheets, BA, RN; Laura
Mueller, BS
Andrus Gerontology Center, University of So. California,
University Park, Los Angeles, CA 90089-0191
Sponsor: Rehabilitation Research and Training Center on Aging with Disabilities, Rancho Los Amigos Medical Center, Downey, CA
PURPOSE--The purpose of this research is to investigate current policies and practices that affect the use of and access to assistive technologies (AT) that support employment and maintain community-based living among adults with disabilities. The goal of the research includes: 1) describing current patterns of AT use and recent changes in patterns of use; 2) examining the impact of AT on health status and on quality of life issues; 3) identifying the barriers, both attitudinal and financial, which may limit access to AT; 4) investigating the adequacy of current policies on AT; and 5) identifying the need, if any, for policy changes.
METHODOLOGY--Both state- and consumer-level policy research are being conducted, drawing upon a variety of methods, including review of existing federal, state, and consumer-level policy reports and research, interviews with key informants, additional analyses of national survey data, and additional project studies to identify consumer issues and investigate state policy efforts.
PROGRESS--We have completed a nationwide survey of state legislative analysts about a variety of state policies affecting AT and a follow-up study of 10 State Units on Aging (SUAs) with high levels of effort in home modifications, and we have conducted analyses of RESNA consumer data on adults age 40 and older to determine particular problem areas, unmet need, types of problems encountered, and receipt of AT information and referral services; of AT data from the 1990 National Health Interview Survey (NHIS) to examine the types and use of AT among older adults and identify barriers to access; and of data from a study of SUA and State Tech Act Projects conducted in 1995 by the American Society on Aging.
RESULTS--The report entitled "Assistive Technology and Adults Aging with Disabilities: A Report on Federal and State Policies and Programs" describes and assesses current knowledge about federal and state policies relevant to the AT needs of middle-aged and older adults with disability; identifies gaps in our knowledge of AT policies and programs; reports findings from additional project studies; and considers what research on AT policy should be conducted in the future.
FUTURE PLANS--Two studies will be completed during 1996/97. First, data collection is currently underway on the collaborative survey entitled "Changing Needs and Life Circumstances of Persons Aging with Disabilities." Our section of the survey focuses on policy relevant issues including information-seeking related to acquisition of AT; use and abandonment of devices; funding patterns; changing needs over time, and unmet needs for AT. Approximately 250 surveys have been completed so far and analysis will begin in fall 1996. Second, the national "Survey of State Rehabilitation Agencies" is in development and data collection will begin in fall 1996. The purpose of this national survey is to identify the extent to which state rehabilitation agencies are knowledgeable about the AT needs of and serve as providers for middle-aged and older adults. AT issues examined include: agency policies, interagency networks, state initiated AT policies and programs, agency decisionmaking, budget allocations affecting provision of AT, and organizational structures.
Go to TOP.
Last revised Wed 05/26/1999
