| The field of assistive robotics has been developing for many years with few major projects and/or products making headline news. With growing applications for assistive robotics and population of elderly/disabled persons, the quality of life provided by insufficient amount of caregivers creates a huge reason for concern. Why don't we have robots to provide assistive roles? Robots have historically been designed to perform either very specialized assistive functionality or to try to do everything. These approaches result in robots that don't do enough or are too complex in functionality or usability for the cost that consumers and/or insurers would have to pay. This dissertation explores design of a cooperative heterogeneous system of robots that can be used to partition the task space to find an optimal level of functionality and cost in order to address customized solutions for people's needs. For the functionality of the robots, we look to some of the standard Activities of Daily Living (ADL) and Instrumental ADL (IADL) as baselines for helping the elderly and disabled population. We identify mobility assistance, vision assistance, home maintenance and community interaction as some high level support tasks for robots in an assistive robot system to take on. Two robots are introduced, one to assist in floor cleaning and another to assist in the mobility, visual and interactive tasks. To support these tasks, we propose the use of open source software, low-cost commercial-off-the-shelf (COTS) sensor components, integration of the robot system into existing home infrastructure where applicable, and cloud computing. Simulations and experimental results are used to validate the claims made in this dissertation. |
