FOCUS:
Robotics · Health Tech · Mental Health
ROLE:
Strategic UX Researcher · Innovation Strategist
DELIVERABLES:
Design Guidance · Adoption Plan
CONTEXT:
In-Home Deployment · Longitudinal Study
Designing In-Home Social Robots for Mental Health Support
A four-week in-home study with older adults experiencing depression revealed key design needs—and demonstrated the measurable impact of socially assistive robots on mental health.
Impact at a Glance
How might we design socially assistive robots and embedded sensing systems to provide meaningful support to older adults with depression, in their actual homes — not just in institutions or labs?
UX Researcher · Innovation Strategist · Design Strategist
I led UX research and stakeholder workshops to drive early alignment around innovation goals. By designing and executing a long-term in-home intervention, I uncovered deep behavioral insights, which I analyzed alongside interview and survey data to identify key patterns. These findings were synthesized into strategic design and product recommendations that informed future development.
This month-long, in-home, proof-of-concept study used a mixed-methods approach to explore the impact of social robots on users’ emotional well-being as well as data for early prediction of depressive episodes. The intervention involved daily robot interactions, embedded sensor logging (via Paro's collar as well as a wearable tracker), weekly structured interviews and mental health (PHQ-9, WHOQOL-BREF, and UCLA loneliness scale) and robot perception (Godspeed, Almere) questionnaires. Additionally, pre- and post-intervention focus groups were conducted with both participants and mental health clinicians. I co-led the fieldwork, focus groups, facilitated participant engagement, and helped translate qualitative and quantitative self-report data into meaningful insights. Our approach enabled both subjective and objective evaluation of how the robot fit into users’ daily routines and emotional support needs.
Strategic Design Insights
Companionship wasn’t enough: Design needed to deliver real support
While users appreciated the robot’s friendly presence, that alone didn’t drive sustained use. They needed help with sleep, security, and emotional regulation — requiring a product strategy that prioritized practical, attuned support over novelty or charm.
Product & Innovation Insights
Sensor data enabled early mental health insight
In-home robots and wearables predicted depression changes with ~74% accuracy, unlocking passive, nonintrusive monitoring.
Prior experience shaped perception
Non–pet owners viewed the robot as more intelligent and lifelike—revealing key opportunities for audience segmentation and expectation-setting.
Reframing as supportive reduced resistance to in-home monitoring
Older adults were surprisingly open to sensor tech when it was positioned as supportive rather than intrusive and was comprised of low-tech sensors.
Clinician feedback shaped strategic viability
Sensor and interaction data were explored with healthcare teams to evaluate real-world usefulness, ensuring the product aligned with clinical care models and workflows.
Companionship — Desirable but Not Sufficient
Pet Non-Owners Perceived the Robot More Favorably
Intervention Reduced Depression & Enabled Depression Prediction
Passive sensing can be more powerful than flashy AI
This project deepened my appreciation for quiet, passive sensing as a clinical tool. Not everything needs to be “smart” in a flashy way to deliver real value in mental health interventions.
Real-world testing challenges lab assumptions
Working in real homes reminded me how often context upends controlled expectations, especially in human-robot interaction. Real-world environments surfaced unexpected contradictions and richer insights.
The richest insights came from cross-disciplinary input
The most strategic breakthroughs didn’t come from any one method. It was the interplay of user interviews, clinician feedback, and behavioral data that shaped truly viable product directions.
Emotional attachment outlasts the product
I was surprised by how many users expressed grief after the robot was removed. This experience reshaped how I think about the ethical deployment of emotionally intelligent technology.
Designing for systems means designing beyond the user
The intersection of robotics, healthcare, and mental wellness required me to think beyond any single discipline. The most effective strategies came from integrating design research, sensor data, and clinical input into a shared solution space.
Validated by clinicians for future trial expansion.
Clinician feedback confirmed both the relevance and feasibility of sensor-enhanced SARs in real-world care, laying the groundwork for scaling to broader trials and adoption pathways.
Informed next-gen SAR design strategy.
Strategic insights, especially around sensor data, user perception, and clinical workflows, directly shaped refinements in both hardware and software direction for future socially assistive robots.
