NRI: Collaborative Research: ASPIRE: Automation Supporting Prolonged Independent Residence for the Elderly

University Of Illinois At Urbana-Champaign, Urbana IL

Investigators

Abstract

Because of the graying of the population, there is a growing need for new assistive technologies to aid the elderly in their daily living. Based on figures provided by the U.S. Census Bureau, and due largely to the aging of the "baby boomer" generation, the population of U.S. adults who are 65 and older is projected to be twice as large in 2030 as it was in 2000, increasing from 35 million to 71.5 million and representing nearly 20 percent of the total U.S. population. This trend is placing enormous burdens on health care costs and causing disruptive changes to how individuals and families manage key late-in-life decisions, including residence. A recent (2015) report by the U.S. Department of Housing and Urban Development concluded that most seniors would prefer to age in place, and a 2010 survey found that 88% of respondents over 65 preferred to remain in their homes as long as possible. It is important to note that these residential preferences must typically be viewed in light of alternatives that are likely to be much more expensive and/or more socially taxing, such as older adults living instead in hospitals, assistive living facilities or with family members. While estimates of the costs associated with these trends vary greatly, it is almost certain that extending the portions of older adults' life spans in which they can live safely and independently through technological means could have enormous positive societal impact. In order to assist in successful aging in place, assistive robots have been developed in the past few decades. However, very few of them have become commercially available, and their use in domesticated environments remains highly limited. The major cause of the problem is that assistive robots typically take the form of full-size humanoid devices or something equally as cumbersome, expensive, and limited in movement and function. We propose a novel assistive robotic system that provides: (i) flexibility, allowing the designed system to be personalized based on users' needs without demanding any home modification upon installation; (ii) safety, ensuring that the system development process accounts for perceived safety by the user, and that the underlying theoretical framework guarantees collision avoidance; (iii) usability, consisting of a minimal and intuitive user interface to provide acceptable controls; (iv) reduced costs, with respect to currently available solutions on the market. The idea behind this research project is the development of a general framework that enables a team of unmanned ground vehicles and small multirotor unmanned aerial vehicles to safely cooperate with the elderly in a home environment. Equipped with appropriate human-machine interfaces, the co-robots will be able to accomplish a number of tasks as demanded by the users. The project addresses fundamental problems in the domain of multi-agent cooperative systems, comprised of humans and co-robots interacting in shared, highly constrained spaces. In order to assist humans, the co-robots have to be trusted by humans, implying that their behaviors be predictable and consistent with principles of human spatial perception, and their appearance must foster a high level of comfort and not create high cognitive demands on the user. Inspired by these challenges, this proposal focuses on the design and control of co-robots, which can adapt to unstructured and rapidly changing environments in a manner consistent with human perception and cognition, thus enhancing safety and robustness. The key focus areas include the design and acceptance of mobile ground and aerial robots that coexist in environments inhabited by humans and the development of a multi-objective control framework to allow intuitive user control over an ensemble of co-robots, which includes the design of both low-level controllers (LLC) and a supervisory, high-level controller (HLC). To demonstrate the benefits of the framework and to engage student groups from various, diverse populations, the following scenario will be considered as a test case: multirotor unmanned aerial vehicles and ground robots acting as domestic assistive devices for healthy older adults in a research laboratory. These co-robots will safely navigate the shared space and accomplish domestic tasks requested by humans while displaying behaviors and appearances that are perceived as safe and trusted. Humans will use an intuitively designed interface for both controlling and monitoring co-robots on a tablet or a smartphone device. A motion capture system and virtual reality Cube at the Beckman Institute will provide the context for data collection, iterative testing and validation.