9561143 Friedman This Small Business Innovations Research Phase I project focuses on research to develop a new and more powerful user interface for advice giving and is built on 10 years experience designing and testing the use of custom wearable talking computers for assisting people with cognitive impairments. Based on an existing wearable computer, a position sensing system will be developed that can be cost effectively installed in schools and potential job sites. The research is to actively sense user location and gross motor activity to give goal oriented advice based on task and user activity models. Use of these models should enable giving advice verbally either after a user request or when there is sensed user difficulty. In our domain of gross motion, we will be able to determine whether or not computer originated advice is heeded. By actively sensing task progress, we will be able to offer encouragement when advice is heeded and provide more detailed guidance when initial cues are insufficient. Finally, we will be able to automatically radio for human supervision if we sense dangerous activities or an inability to follow computer generated guidance. This interface avoids the natural language understanding problems inherent in other active advice giving systems that engage in dialog with their users. It will explore the domain of understanding user needs from task related gross motor activities. The application to enable cognitively impaired users to independently hold jobs rather than remain in custodial care is important to society. If affordable wearable computer technology and minor modifications to a worksite can enable a person to change from being a dissatisfied net consumer of public funding in custodial care to a more satisfied wage earner in competitive employment, it is obviously a win-win-win situation. Society wins by decreasing net care costs to cognitively impaired citizens, the citizens are more satisfied with their lives, and our company has a new, pro prietary product.
