Talking to robots: A research agenda on how to improve the understanding by social-assistive robots of older people
Ronan Flynn, Katrin Paldán
Full text PDF 
( Download count: 12)
AbstractThe world’s population is aging; life expectancy today is much longer than it was fifty years ago. Advances in medicine and improved knowledge regarding healthy living have all contributed to people living longer. Increased human longevity is a cause for celebration but also requires supportive environments that enable older people to maintain a desired level of independence and to manage chronic health conditions, the prevalence of which increases with age. With increased age, there is often a need for care for older (over 65) and very old (over 85) people, be that at home or in health care facilities. This in turn has led to research and development of social-assistive robots, which are designed to interact with older people to provide company and to perform assistive care tasks. The modality most commonly used for people to communicate with assistive robots is speech. However, variations in spoken dialect and the changes in speech articulation due to aging can impact the robot’s speech recognition accuracy and its ability to correctly respond to a spoken command or request. The result of this is frustration when interacting with the assistant robot and, therefore, a lack of acceptance. To improve the interaction with, and the acceptance of, assistant robots by older people, the speech recognition algorithms used must consider the speech characteristics of older people. Following a narrative review of the literature, this paper presents a conceptual framework and a research agenda for more integrated approaches that simultaneously consider both age-related and dialect-related variability in order not to disadvantage older people in speech recognition.Keywords: social-assistive robot, speech recognition, dialect, data augmentation, older people
Ronan Flynn, Katrin Paldán (2026). Talking to robots: A research agenda on how to improve the understanding by social-assistive robots of older people. Gerontechnology, 25(3), 1-14
https://doi.org/10.30289/gt.202604_25(3).1320