Stress Sense
Detecting Stress Before It Strikes
Detecting Stress Before It Strikes
While working with children on the autism spectrum, I began noticing one of the biggest challenges — detecting stress before it turns into visible anxiety or aggressive behavior. That realization sparked the idea for StressSense: a tool that uses wearable data to provide early warnings of emotional stress.
The goal was simple — if we can detect when stress begins to rise, we can act sooner. Over time, this evolved into a broader platform designed not just for ASD support but for anyone who experiences stress. By helping users recognize their physiological triggers early, StressSense encourages conscious behavioral changes and self-regulation — turning awareness into calm, and data into action.
The concept is now patent-pending, with an iOS companion app in development, and will be presented at Apple’s WWDC 2026. I’m currently collaborating with members of the Apple development team to explore integrating StressSense as part of the Apple Watch health suite, expanding its reach to millions of users worldwide.
ROLE & RESPONSIBILITIES:
Developed the idea for StressSense after observing early stress indicators in ASD students. Designed and conducted experiments using Apple Watch HRV and motion data to establish measurable stress-response baselines.
Collected and analyzed multi-sensor data (HRV, accelerometer, and ambient audio) to identify consistent stress signatures. Built an early-stage stress pattern model to differentiate between calm, focused, and high-stress physiological states.
Partnered with mentors in neuroscience and behavioral health to refine the scientific methodology. Led the patent filing process and now working with Apple’s wearable technology team to integrate StressSense into the Apple Watch ecosystem. Currently developing the iOS prototype app, with features for real-time alerts and trend-based stress insights.
Program Highlights & Accomplishments
Patent-Pending Innovation:
Filed a U.S. patent application for StressSense, covering the AI-based detection framework and adaptive thresholding logic used to predict stress onset through HRV, motion, and acoustic data.
iOS & WatchOS Prototype Development:
Designed an early prototype for the StressSense mobile app and Apple Watch integration, featuring real-time HRV tracking, trend visualization, and haptic notifications that alert users when stress levels begin to rise.
Data Science - Neuroscience Integration:
Built a multi-sensor data pipeline using Apple HealthKit APIs and Python-based signal processing to analyze over 25,000+ HRV and motion datapoints, identifying key correlations between physiological variability and self-reported stress. Designed the framework to later integrate EEG and behavioral markers to complement HRV data — laying groundwork for future lab-based validation studies.
Apple Collaboration & WWDC 2026:
Partnering with the Apple wearable technology team to refine app design, optimize battery efficiency for continuous monitoring, and prepare for a live demonstration at Apple’s WWDC 2026, where StressSense will be showcased as a next-generation emotional health feature for the Apple Watch suite.
Community Impact Vision:
Extending the tool’s use beyond research — creating a companion caregiver dashboard for parents and therapists working with neurodiverse children to receive alerts, observe stress trends, and apply personalized calming interventions.
Vision Ahead
To evolve StressSense into a multi-sensor predictive platform that integrates HRV, EEG, and behavioral data for proactive emotional-health monitoring.
The vision is to make StressSense a core component of Apple Watch Health, offering real-time alerts, personalized feedback, and long-term emotional-wellness insights — helping users recognize stress before it strikes and choose calm, consciously.
Feedback
“He’s not just collecting data; he’s learning to identify meaningful patterns and understand the small signals of the body.”
- Sankar Dasiga, Mentor
Ongoing Journey
It all started while working with kids on the autism spectrum. I noticed how hard it was to tell when someone was getting stressed — things only became obvious when they were already overwhelmed. That’s when I thought, what if a watch could tell us earlier? .
I started diving into how these gadgets actually work, learning about different sensors, signal types, and measurements, and figuring out which ones really matter for detecting stress. Every test made me better at understanding how to turn raw data into something meaningful.
I started talking to neuroscience mentors who helped me understand the science behind stress — not just the data. That’s when StressSense shifted from being a “cool idea” to something that could actually help people.
After months of testing, the data finally started showing patterns. I could actually see when someone’s stress level was rising before they felt it. That moment changed everything — I knew it could work.
Learning everything about patent filing — from prior art searches to claims and provisional applications. It’s complex (and honestly, a bit overwhelming at times), but also incredibly exciting to see my work move from an idea to something officially protected. Every new section I read feels like unlocking another piece of how innovation becomes real.
When I first started doing the POC for StressSense, I noticed that the app stopped collecting data after about 15 minutes when running in the background. To troubleshoot, I switched it to active mode, which allowed me to monitor data continuously for 60 minutes and better assess how the model performed in real time.
The next challenge came with Learning Mode, which required longer, uninterrupted data sessions to build a user’s baseline. Initially, I wasn’t sure how to keep collecting data while the app wasn’t open — but then I discovered that I could leverage Apple HealthKit’s background data sync. By running a sync process whenever the user opens the app, StressSense can now pull historical HRV and activity data collected in the background to update and refine the baseline model.
This was a big step forward — turning what seemed like a technical limitation into a smarter, more user-friendly solution for passive data collection and long-term stress tracking.