DESIGN + ENGINEERING + BEHAVIORAL SCIENCE + DATA
The World Health Organization defines health as “a state of complete physical, mental and social wellbeing and not just the absence of infirmity.” We combine design, engineering, and behavioral science principles to create interfaces and data-intensive systems to keep people healthy and performing at their best. We aim at creating individualized and adaptive behavior change technology interventions seamlessly embedded in daily life routines. In short, we want to find people where they are (e.g. sitting, commuting, talking, writing, chatting) to help them achieve wellbeing through life.
AREAS OF STUDY
We focus on four areas of study, two core research areas, an exploratory one, and a collaborative one. Core research on Passive "Sensorless" Sensing and Subtle Just-in-Time Interventions leverage human-computer interaction (HCI), data science, and human-centered artificial intelligence (HAI) to design creative solutions based on scientific foundations to promote wellbeing. We also explore topics related to Shared Autonomy for Behavior Change through novel ideas combining behavioral science, peripheral interventions to propose new behavior change theory. Finally, we collaborate with colleagues from Chemical and Electrical Engineering to create ecologies of devices to detect Precise Mental States combining bioelectric, biomechanical and biochemical sensors for continuous longitudinal stress monitoring.
We design just-in-time interventions that balance efficacy and engagement for everyday stress management to promote wellbeing and mental health when needed. We modify cyber and physical spaces to deliver micro or subtle interventions with minimal to no distraction to the user.
We propose using our browser-based and mobile Stanford Wellbeing and Emotion Education Technology (SWEET) platform to deliver support for people working from home (Home Sweet Office), or studying from home (Home Sweet School), especially now during COVID-19 times.
In the past we have developed projects such as the Popbots, where we used an "army" of tiny chatbots for stress regulation, or the Mindful Commute, where we leveraged guided breathing regulation in the car.
Passive "sensorless" ensing is an adaptive approach that repurposes data from existing devices or data extracted from existing sensors embedded in everyday objects to passively measure stress, and performance.
We propose AI-enabled multimodal stress sensing by combining biochemical (cortisol) with biomechanical passive stress sensors using data from PC mice or trackpads to detect stress in multiple populations (office and clinical workers). In our project Fast & Furious, we also leverage the car steering wheel as a valid stress detection device.
In the past we have developed seminal work on biomechanical stress, such as detecting acute stress with the mouse (MouStress) or trackpad (Stress Tracker).
These are our forward thinking theoretical ideas on how "invisible" peripheral technology can enable new frontiers of research in behavioral sciences. We explore automation of everyday devices to enable non-volitional behavior change or subliminal haptic entrainment to regulate behavior through automatice regulation of mental states by interfacing with human's interoceptive processes.
We led a group of researchers to win a the prestigious NSF NSF Multimodal Sensor Systems for Precision Health Enabled by Data Harnessing, Artificial Intelligence, and Learning (SenSE) program with our proposal: Artificial Intelligence-enabled Multimodal Stress Sensing for Precision Health. We motivated Prof. Zhenan Bao and Prof. Mert Pilanci to join us to make a proposal to combine biochemical, physiological, and biomechanical sensors and data to increase mental health measurements.