September 16, 2021
Project EMAR works to improve adolescent mental health via a co-designed social robot
According to data from the National Institute of Mental Health, about 3.2 million teens in the United States have had at least one major depressive episode within the past 12 months.
These findings confirm that depression and anxiety among teenagers have been on the rise, which provokes questions around whether current school-based mental health services sufficiently support teens.
To address this mental health crisis, Elin A. Björling, senior research scientist and part-time lecturer in the Department of Human Centered Design & Engineering, has worked to create EMAR, an innovative social robot.
“Stress is the ubiquitous exacerbator,” Björling said. “[Teenage years] are when stress really starts to appear and becomes difficult to manage. That’s the time point that an intervention could potentially change the trajectory of adulthood.”
To better understand the target population, Björling began examining local school systems in Seattle. She soon realized there was a shortage of individualized resources for teen mental health.
Existing resources usually require direct interaction with parents and school staff, which may be daunting and deter teens from utilizing the available resources. Offering teens more agency in controlling their mental health was a critical component that was missing.
“I realized we needed to think out of the box and I wanted to start thinking about community-based resources,” Björling said. “I was doing consulting with different research teams and learned from colleagues who were using design thinking and human centered design. That’s how the wild idea of robots got on my radar.”
A social robot dedicated to mental health would help reduce the existing barriers to seeking mental health treatment and provide an accessible micro-mental health intervention for teens in an academic setting.
“I thought a robot could be really engaging for teens,” Björling said. “It could allow teens to offload their troubles without human judgment, acting as a soft landing and community-based tool for starting a conversation [about mental health].”
Björling and an interdisciplinary team of researchers began their project by first analyzing their target population, observing and collaborating with teens to co-design what would become the Ecological Momentary Assessment Robot, also known as EMAR.
From 2016 to 2020, the researchers worked directly with teens to understand teen mental health and needs. The researchers also interacted with the teens to help design the robot, including special features and characteristics intended to be user-friendly.
“For teenagers especially, [human centered design] makes so much sense,” Björling said. “Having them engaged in the process of the design…is the only way to ensure, at the very least, that you are making something that the user actually wants to participate in.”
The robot was designed to conduct ecological momentary assessments, a way to sample teens’ moods and experiences in a natural environment in real-time.
“Often in health spaces, we use retrospective assessments, which is not always reliable since humans are not always accurate in tracking their experience,” Björling said. “During my doctoral program, I learned about ecological momentary assessment, which is an approach … that evaluates people in the moment to ensure accurate and reliable data.”
This type of assessment enables the researchers to collect reliable data on teen mental health, providing an empirical appraisal of teen mental health outcomes over time. Administering these assessments through the social robot is intended to offer teens an engaging and less intimidating tool designed to provide mental health support.
Among the design features that teens suggested, the first was ensuring the robot interaction was anonymous. In addition, teens wanted a tabletop device that expressed empathy through sounds, voice, and movement. This design affords teens greater discretion to independently approach the robot instead of feeling uncomfortable with a robot that might approach them.
“The teens wanted to approach the robot, interacting with it on their terms,” Björling said. “We also had data from teens wanting to hug or interact with the robot in a way that physically communicates emotions.”
To accommodate this desire, the researchers began working on programming EMAR to tilt its head, to signal empathy and attentive listening.
Project EMAR received funding from the National Science Foundation in 2017, but the team was struggling with developing the robotic core. After thoroughly conceptualizing EMAR’s new core, Björling’s project received funding from a 2020 Population Health Innovation Award grant from CoMotion and the Population Health Initiative.
Björling’s project, alongside Kit Galvin’s PestiSeguro/PestiSafe app, was awarded funds due to the project’s potential for improving population health and achieving sustainable economic or societal impact.
“The award allowed us to hire mechatronics designers to help solve the mechanical problem we ran into of getting the robot to tilt its head, which freed more time to do greater discovery and assembly work,” Björling said.
Additionally, Project EMAR was selected as one of three projects featured in the summer 2019 Social Entrepreneurship Fellows program, which is facilitated through a collaboration between the Population Health Initiative, the Buerk Center for Entrepreneurship, the Evans School of Public Policy & Governance and CoMotion.
“The support we received through the fellowship program gave us a tactical understanding of the mental health resources and needs for several schools across the country,” Björling said. “We realized that many schools lack money for mental health but have funding for STEM-related programs. This got us thinking that if [EMAR] had some STEM educational components, it could be a perfect fit as both a mental health tool and tool for STEM literacy.”
The researchers are now gearing to deploy the system as a device customized, programmed, and operated by teenagers. The system will be provided as a mental health tool and a valuable mechanism to improve STEM literacy.
“One of the biggest findings was the diversity and customization students wanted for the robots,” Björling said. “We realized that for this system to work in different high schools, with different populations, we needed to make the system highly customizable.”
The team worked hard to create a flexible system to ensure that each community receiving an EMAR could individually code the robot and customize its body to accommodate the school’s needs.
“Having teenagers interacting and coding with the robot enables it to be a sustainable and affordable educational robot,” Björling says. “This allows the system to not only address teen mental health as a community tool but also as an educational resource.”
As EMAR continues to evolve and be pilot tested, the research team notes the potential for their system to be expanded to a broader audience of students from different grade levels and becoming an embedded part of a community system.
“There is an enormous potential impact, through getting [teen mental health] data out and into a community and translating simple mental health tools onto a robot interaction. EMAR will provide a practice space and engage teens in mental health exercises in a fun way,” Björling said. “Who doesn’t want to have a quick conversation with a cute robot?”