I attended the American Association for Artificial Intelligence (AAAI) Fall Symposium on Developing Assistive Technology for People with Disabilities at MIT, Cambridge, MA (November 9-11, 1996). This symposium was the result of an increasing interest in the mainstream AI community towards research focusing on the needs of people with disabilities. It follows last year's workshop on "Developing AI Applications for People with Disabilities" held in conjunction with the 1995 International Joint Conference on Artificial Intelligence (IJCAI-95), Montreal, Canada. This increased interest was highlighted by Dr. Raj Reddy's Turing Award acceptance speech where he identified the application of AI theory and technology to the needs of people with disabilities as one of the most promising areas in which to work-both in terms of results as well as impact on society.
This symposium was intended to bring together AI researchers to focus on open problems and issues as well as discuss where work in assistive technology should be heading. Discussions also included technology transfer to and from the AI community and convincing the AI community that this is important research.
The papers in the symposium were broadly categorized into three categories:
The keynote speaker was Paul Meyer of the President's Committee on the Employment of People with Disabilities.
He identified four issues:
He was more than willing to solicit ideas and feedback from people. To participate, contact Paul Meyer, President's Committee on the Employment of People with Disabilities (phone) 202-376-6200 (fax) 202-376-6219 (e-mail) pmeyer@ pcedpd.gov
After each set of related presentations, the attendees broke into working groups to discuss open problems and issues. While time and space constraints prevent me from elaborating on each of the issues we addressed, I would like to highlight some of the major concerns.
One of the problems in getting more involvement by the AI community in working in assistive technology has been the lack of acceptance by review committees of papers that are considered purely application oriented. We discussed how to bring about a greater acceptance of applied research by the "mainstream" AI community. We agreed that papers and results should explicitly spell out what advancement the specific application is addressing, and should also spell out generalizations that can be applied across the board. The working group also addressed different ways to increase the awareness of the AI community that assistive technology research can be the perfect and honorable validation of research. We discussed whether it would be useful to consider a special session on assistive technology as part of the regular AAAI meetings but decided that it would not conducive towards our goal since it will compartmentalize researchers working in the field. We agreed to look into starting a chair for assistive technology in the AAAI fold, having an invited speaker in a general session, and holding a special plenary session on assistive technology.
It was noted that both evolutionary and revolutionary research has its place. While incremental advances are good and useful for immediate benefits, we should also not abandon basic research.
This formed a major part of our discussion. We explored different ways to try to increase the funding for assistive technology research by such organizations such as NIH and NSF. With this goal in mind, we decided it would be useful for the AAAI executive body to approach these agencies and lobby for increased support in assistive technology research.
It became apparent (based on some of the papers presented) that the rapidly aging population, both in the West and in Japan, is bringing in a sense of urgency in developing assistive technology. We were presented with some statistics that highlight this fact: 25% of Europeans will be over 65 years old by the year 2020. Of the 1.5 million residents of US residential care facilities, 70.7% over the age 65 had some sort of mobility impairment, and 22.7% had some visual impairment. Among those over 85 the proportion of visually impaired increased to 30%.
While there may have been general assent about involving consumers right at the beginning, it appears that most of the projects currently have little or no input from consumers until the end of the work (if at all). Also, there was little discussion on the actual role consumers should play and to what extent. However, it was mentioned that in Europe, disability groups play a much more direct role in deciding what research actually gets funded (and what the results should be).
The existence of a communications standard in Europe (M3S) for assistive tech components to talk with each other was mentioned and a brief discussion was started. Also, several people brought up the need for standard and appropriate text corpora to be used in the evaluation of language aids. --Zunaid Kazi, Applied Science & Engineering Laboratories
URL of this document: http://www.asel.udel.edu/
Last updated: February 6, 1997
Copyright © Applied Science and Engineering Laboratories, 1997.