Close

Presentation

Exploring the potential of the Stretch mobile robot for older adults with mobility disabilities
Event Type
Oral Presentations
TimeWednesday, April 1411:40am - 12:00pm EDT
LocationDigital Health
DescriptionOlder adults desire to successfully age in place, defined as the ability to live independently in one’s place of residence and maintain social engagement in their community (Rogers, Ramadhani, & Harris, 2020). However, older adults with mobility disabilities, the most common disability among older adults (US Census Bureau, 2014), are especially at risk of experiencing declines in autonomy, disease, hospitalizations, and poorer quality of life (NIA, 2020). These limitations increase barriers to aging in place (Haak et al., 2007). In particular, older adults aging with mobility disabilities commonly require support with instrumental activities of daily living (IADLs), such as cleaning, meal preparation, medication management, and housekeeping, which must be performed to successfully age in place (Wiczorek et al., 2020).

Assistive mobile robots hold great promise for promoting independence, home healthcare, and aging in place for older adults with mobility disabilities (Haak et al., 2007). Yet, assistive robots are not widely commercially available (Wiczorek et al., 2020) perhaps due to their size and cost (e.g., PR2, Beer et al., 2012). A novel mobile and assistive robot developed by Hello RobotTM, StretchTM, is a first-generation research edition mobile platform that is compact, relatively inexpensive compared to other mobile manipulators, and capable of performing IADL tasks by using a telescoping arm and compliant end-effector grippers. Stretch is currently available for researchers to explore how robots might assist with IADL tasks.

Stretch was designed to support individuals with disabilities. The goal for assistive robots would be to have them in people’s homes to provide the support that they need. To achieve this goal, human factors evaluations are needed from initial deployment to long-term use in homes. As a first step to evaluating this potential, we evaluated the initial unboxing and setup of the Stretch RE1. We used subject matter experts (SMEs) and human factors principles to consider obstacles older adult users might encounter during the unboxing process. We then evaluated 32 trials of user testing and critical task analyses in a home-like environment to ensure usability and ease of use. All study procedures took place in the McKechnie Family LIFE Home (Living in Interactive Future Environments), which is a smart home research facility on the University of Illinois Urbana-Champaign campus. Research team members controlled Stretch by using an Xbox controller that was programmed to operate Stretch. We used Stretch to manipulate 13 different types of objects that would be part of IADL tasks, such as used in meal preparation. We documented the complexity of subtasks, types of errors, time-spent manipulating the object, and informal qualitative feedback from tele-operators during and after each trial (using a think-aloud protocol). Our goal was to identify barriers to use by older adults with mobility disabilities, as well as design adaptations (e.g., new end effectors) for the robot that could support IADL tasks.

Our review of the unboxing process revealed potential physical access barriers as well as usability issues with the instructional support materials. The results from our critical task analyses showed that, on average, more errors occurred with environmental manipulation (e.g., opening cabinet doors) compared to object manipulation (e.g., moving kitchen utensils into the sink). Objects that rotate about a fixed axis were the most difficult to manipulate. For example, vertical cabinet handles and microwave doors posed the most challenges for users. Inspired by the design of prosthetic hooks commonly used to support individuals who experience upper body mobility disabilities with this type of task, we developed a novel end-effector hook with a modular design to work with existing Stretch parts (e.g., standoffs and screws) to enhance the usability of the robot (e.g., fewer errors). Our iteratively designed end-effector hook yielded successful completion of the cabinet task by multiple tele-operators and resulted in fewer errors compared to the initial compliant end-effector and Puller trials.

In summation, assistive mobile robots are emerging technology and not yet widely commercially available nor widely adopted. However, the potential for mobile robots, such as Stretch, to promote successful aging in place for older adults with mobility disabilities was revealed by our critical task analyses and testing in the LIFE Home. Our findings highlighted potential usability limitations and challenges that older adults aging with mobility disability may encounter when unboxing and operating Stretch to do IADL tasks. Our approach is the first step in a human factors evaluation to advance the design and usability of assistive mobile robots. Next steps will be to engage older adults with mobility disabilities through the next iterative robot design cycles. By directly including older adults as co-designers, mobile robots can be tailored to their specific health needs and preferences.

Potential Conflict of Interest: Authors Edsinger and Kemp are the owners of HelloRobot, which produces the Stretch robot. All other authors have no conflict of interest to report.