Robots and Avatars that Show Emotions and the Role of Telepresence during the COVID-19 Pandemic
by Elena Portacolone, PhD on 15 June 2020
In the COVID-19 pandemic, in the absence of a cure or vaccine, the social distancing of older adults with cognitive impairment is vital for their survival. During the pandemic, to protect their residents, most institutions (such as hospitals, nursing homes, or assistive living facilities) have denied access to family members unless the individual is actively dying . Adult day health centers and senior centers have closed. Older adults who have cognitive impairment and who live in communities are sheltered in place in their own homes , and many of them live alone. A recent study of our group estimated that, in the United States, about 4.3 million adults whose ages are 55 and over have cognitive impairment and live alone , and this number of people is greater than the entire population of the city of Los Angeles.
Yet, this imposed physical isolation is likely to increase the unmet needs, as well as the loneliness of older adults with cognitive impairment. Telepresence applications, such as the pet avatar discussed in our case study , make it possible for these adults to have connections with other people, e.g., medical providers, physical therapists, family members, and friends, for instrumental, informational, and emotional support. Thus, the COVID-19 pandemic is showing us the urgency of interdisciplinary collaborations that ensure that emerging technologies, including the pet-avatar, clearly serve and protect users’ values and lives. These collaborations could also ensure that these technologies are given to those who need them the most, especially during emergencies, such as the current COVID-19 pandemic. For example, priority should be given to older residents of neighborhoods with high concentration of cases of COVID-19 and older adults from racial/ethnic minorities because they have been particularly affected by the virus [5, 6].
In this pandemic, technologies such as those involving telepresence can provide solutions. For example, the immunity of machines to the coronavirus is allowing access and activities in places where people’s health would be at risk. In some airports, robots are detecting passengers with symptoms of the virus . Similarly, robots could detect the symptoms of the virus in users with cognitive impairment and their visitors. Robots also are disinfecting hospital rooms and aircraft cabins by means of the emission of ultraviolet light . Maybe this same process could be used to disinfect the homes and deliveries of older adults who are at risk. Also, robots are delivering food and medical supplies . For example, older adults could ask virtual digital applications like Siri or Alexa to make a grocery list and have it delivered by a local supermarket. Once again, applications such as the pet-avatar can be useful to organize home deliveries, especially considering that some countries already have robust systems of home deliveries. The opportunities are endless.
In our Ethics Review , published in the Journal of Alzheimer’s Disease, we provide a case study describing a software, Care Coach, that involves an avatar that provides company and support to older adults with cognitive impairment and which is controlled remotely by technicians. A second paper by Robillard et al.  describes some of the broader ethical themes relevant to this case, and social robots in general. The third paper by Robillard et al.  discusses how to address a problem of underutilization of social robots in general by proposing a collaboration or cocreation of devices by involving end users early in the development process to ensure the end products are consistent with needs and values.
Prompted by the second paper by Robillard et al. , we are taking this opportunity to expand on our suggestion that the pet-avatar should look like a robot rather than a pet and to limit the avatar to forms of support that do not include expressions of affection.
One of the key ethical concerns we raised in the case study is that some users with cognitive impairment may come to believe that the pet avatar is real. Specifically, presenting the avatar as a pet could cause patients to infer that this creature possesses many of the aspects typical of a pet such as unconditional love and dependence on them for care. Furthermore, families may feel compelled to continue the paid service because they might fear patients may suffer needlessly if they experience the sense of loss a pet owner might feel. Changing the avatar to a robot rather than a pet and limiting the avatar to forms of support that do not include expressions of affection could be one solution. Our position stemmed from our desire to minimize the opportunities for users with cognitive impairment to think that they are interacting with a pet rather than a machine. Specifically, we contend that we owe it to users with cognitive impairment to consider that their impairment might hamper their ability to understand that there is a human being behind the avatar. Thus, to provide users with cognitive impairment the ability to make accurate choices, developers should focus on providing clear messages and reminding these users of their options. That is why we suggested that the application should be presented as a technology rather than as a pet. In addition, to elaborate on our position, we suggested that the technicians should be visible to users, which would address any concerns about the users being deceived. Also, to avoid deception, we owe it to people with cognitive impairment to avoid using features that are likely result in misinterpretations, such as robots displaying signs of affection in order to be adopted. However, it is useful for robots to be able to discern emotions from human features to detect early signs of depression and to prevent emergencies . As Robillard et al.  emphasized as well in their commentary, we look forward to discussing the endless opportunities offered by telepresence in interdisciplinary gatherings attended by users, caregivers, ethicists, social scientists, developers, and healthcare providers.
Elena Portacolone1*, Jodi Halpern2, Jay Luxenberg3, Krista L. Harrison4,5, Kenneth E. Covinsky4
1Institute for Health & Aging, University of California San Francisco, San Francisco, CA, USA 2School of Public Health, University of California Berkeley, Berkeley, CA, USA 3On Lok Lifeways, San Francisco, CA, USA 4Division of Geriatric Medicine. University of California San Francisco, San Francisco, CA, USA 5Philip R. Lee Institute for Health Policy Studies, University of California San Francisco, San Francisco, CA, USA
Correspondence to: Elena Portacolone, MPH, MBA, PhD, Associate Professor, Institute for Health & Aging, University of California San Francisco, 3333 California Street, Suite 340, Box 0646, San Francisco, CA 94118, USA. E-mail: firstname.lastname@example.org
ACKNOWLEDGMENTS This work was supported by the Career Development Award (K01 AG049102 PI Portacolone) from the National Institute on Aging, National Institutes of Health. This work was also supported by the New Investigator Research Grant Award (NIRG 15-362325 PI Portacolone) from the Alzheimer’s Association and by the Pepper Center at UCSF (P30AG044281 PI Covinsky), which promotes promising new research aimed at better understanding and addressing late-life disability in vulnerable populations. No funding source had any role in the study design; collection, analysis, or interpretation of data; writing of the report; or the decision to submit the article for publication. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health, the Alzheimer’s Association, or the University of California.
CONFLICT OF INTEREST The authors have no conflict of interest to report.