ExpiWell’s Researcher in Focus: Dr. Casey Bennett on How Socially Assistive Robots (SARs) Can Transform Chronic Healthcare Management

ExpiWell’s Researcher in Focus: Dr. Casey Bennett on How Socially Assistive Robots (SARs) Can Transform Chronic Healthcare Management

Angelo Yanga
Line divider

About The Study 

Integrating socially assistive robots (SARs) presents a promising avenue for managing chronic conditions effectively. However, designing interactions that resonate meaningfully with end-users is a critical challenge. While it's easy for designers to conceptualize interactions, understanding how users engage with these agents autonomously remains complex.

Hence, Dr. Casey Bennett and colleagues' study delves into the role of SARs in chronic healthcare management, exploring methodologies to enhance user engagement and effectiveness. By deploying SARs equipped with advanced sensor arrays in users' homes, the research aims to collect valuable data on interaction patterns and environmental dynamics.

Motivation Behind the Study 

Dr. Casey Bennett shared that the aging population in many countries motivated them to study SARs and their benefits for providing social support. 

He added, "Currently, there is this issue of the "Silver Tsunami" where populations in many countries are rapidly aging, i.e., there are many older people but not enough young people to care for them.  The idea with SARs is they can help fill that gap by providing older adults with social support that improves their health outcomes.  Much like a real-life pet, except that SARs don't need to be fed, and they won't poop on your floor, so they are easier to live with.  Not to mention, many facilities like nursing homes or retirement centers don't allow real-life pets."

According to Dr. Bennett, young adults can also benefit from integrating SARs into their daily lives. "Beyond older adults, we have also found there may be uses for SARs with younger adults, such as those with autism or college students with mental health issues like anxiety.  You can't have a real-life pet in college dorm rooms, for instance, so SARs are a good solution,"

Role of SARs in Chronic Healthcare Management

Chronic health issues, such as Alzheimer's, autism, or diabetes, have no cures, and they can affect the patient's daily lives, which may put them in need of assistance. The role of SARs in chronic healthcare management is to ensure that the patients have enough support in their everyday routines. 

Dr. Bennett also shared, "We mainly work on chronic health issues, which have no cure.  Things like Alzheimer's, Autism, or Diabetes. So the problem is different from acute issues, like a heart attack or cancer, in that the goal is not to help the person recover but rather maintain their health status as high as possible for as long as possible."

In the human-robot interaction (HRI) context, SARs are equipped with sophisticated sensor arrays that collect data about user interactions and their surrounding environment. This sensor data is invaluable for understanding real-time interaction behaviors and can be leveraged to develop models for intelligent robot control.

He added, "Many people don't realize that chronic health issues make up about 90% of healthcare, so they are critically important, and SARs can be helpful there." 

So, the role of SARs in chronic healthcare management helps patients and healthcare professionals by assisting them in caring for patients and gaining more insights into patient experience. 

The Challenges of Using SARs 

Despite their assistive qualities in chronic healthcare management, using SARs can also be challenging. For this reason, the study also explores the interaction between humans and robots. 

Integrating socially assistive robots (SARs) into everyday living spaces presents many challenges, particularly in understanding user interactions across diverse geographic locations. 

This research addresses these challenges through a long-term deployment of SAR companion pets in user homes across South Korea and the United States (23 participants: 12 Korean, 11 US), coupled with sophisticated sampling techniques to generate a large-scale dataset of naturalistic human interactions with the robots.

One of the primary challenges encountered was technical hardware failures, which led to partial data loss and necessitated the exclusion of some participants from the analysis. Despite this setback, the research yielded valuable insights into the nuances of human-robot interactions across different cultural contexts.  

Other challenges may arise from using SARs, such as sporadic engagement, ethical considerations, integration into healthcare ecosystems, cultural variability, and developing meaningful interactions. 

Addressing the Challenges 

Interdisciplinary collaboration and innovative research methodologies are essential to overcome these challenges and maximize the benefits of SARs in HRI. Long-term deployments of SARs in diverse settings, coupled with sophisticated sampling techniques and rigorous data analysis, can provide valuable insights into user behaviors and preferences.

Furthermore, ongoing research efforts should focus on developing culturally sensitive SARs, enhancing technical robustness, and addressing ethical concerns. By effectively navigating these challenges, we can unlock the full potential of SARs in improving healthcare outcomes, enhancing quality of life, and fostering meaningful human-robot interactions.

Methodology & Surprising Findings 

The study was conducted over one year, from August 2021 to July 2022, employing a convergent parallel mixed method approach to gather comprehensive insights into human-robot interaction (HRI) with a socially assistive robot (SAR) companion pet. Here's an overview of the methodology employed:

  • Twenty-six participants, consisting of 13 Korean and 13 US individuals, were recruited for the study. (Due to technical hardware failures during deployment, three participants were excluded from the analysis) 
  • Integration of ecological momentary assessment (EMA) data and Sensor collar data, serving as features, was collected roughly nine times per second, every minute of every day, resulting in approximately 11.7 million data points per participant and a total sensor dataset of over 270 million data points.
  • The study utilized a convergent parallel mixed method approach, incorporating quantitative and qualitative data collection methods. (quantitative - processing sensor collar data and qualitative -  follow-up interviews to gather participants' subjective experiences) 

In summary, the study's results concluded that the survey of SAR deployment in homes across the United States and South Korea uncovered unexpected insights into ML (machine learning) and DL (deep learning) model adaptability. While models performed well in their native location, their accuracy significantly dropped when transferred elsewhere. 

Surprisingly, models trained on combined datasets showed intermediate results. Further analysis revealed cultural and environmental differences, stressing the need to consider social contexts for SAR deployment in healthcare. These findings highlight the challenges of designing SARs for diverse settings and call for further research in the field.

ExpiWell Experience with Dr. Casey Bennett

Real-time data collection is essential for the healthcare industry to give healthcare professionals insights into patient experience.  Such real-time data is the primary goal of EMA. For this reason, healthcare researchers need should use an effective tools for a more successful EMA or ecological momentary assessment study process. 

Like Dr. Bennett and his colleagues, using ExpiWell for EMA has helped them gather information about the use of SARs that can benefit the healthcare industry in the future. 

He also said that "EMA has wide applicability to many health issues and digital health technologies. So I think tools like Expiwell are really useful for researchers who want to engage in such research, especially when it comes to understanding how human health is affected by the choices we make every day beyond the few times a year we show up at the doctor's office.

(Dr. Casey Bennett, Asst. Professor & Chair of Health Informatics at DePaul University)

Future Directions for SAR Development with EMA Method

With the help of EMA and an effective research tool, the future direction of SAR development can be positive and groundbreaking. Although there might be challenges in implementing robotic assistance, using an advanced machinery process can lead to a more efficient healthcare process. 

Dr. Bennett also added, "a big problem with using robots for healthcare, such as SARs, in in-home settings beyond the clinic is that we don't know what is happening with robots or their health daily, in real-time.  That is a problem with healthcare in general.  Traditionally, the only way to gather that information was to ask people what happened later or to keep some "diary," but those methods are full of recall bias."  

EMA is indeed helpful if researchers want to minimize biases in their participant's answers. By minimizing bias, researchers can get a more accurate picture of human health, which can then lead to the development of SARs. 

Finally, he explained that "many people struggle to remember what they ate for lunch yesterday, let alone detailed health fluctuations or how they interacted with the robot over the past few days.  So we have been using Expiwell to do what is known as "Ecological Momentary Assessment" (EMA), combining smartphones and sensor data from SARs to track their health status and health-related behaviors in real-time.  That allows us to gather more accurate data and avoid issues like recall bias."

Invitation to the Research Community

The ExpiWell team is excited to have helped and facilitated research by Dr. Casey Bennett and colleagues. We continue to work toward innovating and enhancing scientific discovery.

We invite you to explore our Journal Publications section for a deeper dive into a range of insightful research studies and to discover how ExpiWell has facilitated critical experience sampling and ecological momentary assessment data collection. 

If you're interested in harnessing the power of ExpiWell, don't hesitate to contact us with any queries or for support. You can contact us or email sales@expiwell.com to learn more about our platform and app. 

Reference: 

  1. Bennett CC, Šabanović S, Stanojević C, Henkel Z, Kim S, Lee J, et al. Enabling robotic pets to autonomously adapt their own behaviors to enhance therapeutic effects: a data-driven approach. IEEE International Conference on Robot and Human Interactive Communication (RoMAN) (2023).

Recent Blogs

Release notes: Newly enhanced features (to help improve the work you do)

October 26, 2021

Release notes: Newly enhanced features (to help improve the work you do)
Read More
Everything You Need to Know About Geographic Ecological Momentary Assessment (GEMA)

November 20, 2024

Everything You Need to Know About Geographic Ecological Momentary Assessment (GEMA)
Read More
ExpiWell Researcher in Focus: Abbey Salvas on How Authentic Expressions Affect TNB Individuals in the Workplace

November 1, 2024

ExpiWell Researcher in Focus: Abbey Salvas on How Authentic Expressions Affect TNB Individuals in the Workplace
Read More
By clicking “Accept All Cookies”, you agree to the storing of cookies on your device to enhance site navigation, analyze site usage, and assist in our marketing efforts. View our Privacy Policy for more information.