User-Centered Hardware and Software Interaction Design Approach to Enhance Accessibility and Utility of Embodied Social Robots for Older Populations
Presented at the Applied Human Factors and Ergonomics (AHFE) Conference, Virtual (July 2020)
Rapid developments in artificial intelligence, machine learning, voice interfaces, and robotics, have enabled consumer social and entertainment robots to flood the market, with sales forecast to reach 4.22 million units by 2025. Notably, embodied social robots have unique advantages over more mature, non-embodied simulated agents such as smart speaker technologies (e.g., Google Home, Amazon Alexa), as they provide an added layer of character (e.g., through humanoid form factor, proactively engaging movements, body language paired with speech, non-verbal communication, user attention), and resultant empathy.
These social robots are entering the market in all shapes and sizes, ranging from tabletop robots, to mobile robots, to humanoid robots, to animal robots, to robots for service industries, robots for security, and robots for healthcare. Although growth in this market is starting to bring more niche-focused robots (e.g., Luka for reading to children), a majority of social robots are designed as general family robots, acting as both companions and assistants to each member of the family. These social robots companies have taken a one-size-fits-all approach to their social robot product offerings, and as a result, these robots have failed to reach the market on time (e.g., Buddy, Zenbo) or have already abandoned development (e.g., Kuri, Jibo). The technology is not the limitation–it is over-promise and ill-defined scope coming at the expense of the specific usability, accessibility, and acceptability needs of a target population. For example, the majority of these robots function entirely though voice-based interaction (requiring volume, articulation, and memory for the appropriate wake word) and/or onboard screen-based interaction (requiring skin conductivity, fine motor control, and forceful touch), limiting suitability for older adult populations with cognitive impairments, such as individuals with Alzheimer’s disease (AD) and AD-related dementias (ADRD), despite the significant potential impact social robots are positioned to provide to these populations.
Charles River Analytics is working on a research and development effort funded by the NIH’s National Institute of Aging (NIA), to adapt and/or augment existing social robots to fit the specific usability, accessibility, and acceptability needs of individuals with early to middle stage AD/ADRD, while also harnessing the unique empathetic qualities of social robots to provide targeted utility to increase social connectedness and reduce loneliness and social isolation for both individuals with AD/ADRD and their caregivers. Paramount to this research effort is a co-creation approach, driving design, features, and functionality through a series of focus groups and interviews with individuals with early to middle stage AD/ADRD and their caregivers, to understand their pain points and build something in which all parties find value. Based on user feedback and iterative user testing, we are designing, prototyping, and augmenting the existing QTrobot with additional hardware and software to enable seamless interaction and improve quality of life. We discussed our high-level findings from our user research sessions, recommendations for hardware and software design to appropriately address the needs of older adult AD/ADRD and caregiver populations, and discussed how our work can be generalized to other targeted populations, as well as future research directions.
For More InformationTo learn more or request a copy of a paper (if available), contact Caroline Kingsley.
(Please include your name, address, organization, and the paper reference. Requests without this information will not be honored.)