CalmBand

• 

  Calmband hardware Architecture

• 

  Basic Architecture

Inspiration Children with autism and Down syndrome often experience sudden sensory overload or anxiety that can quickly escalate into meltdowns. From conversations with parents and educators, it became clear that caregivers usually recognize distress only after it has already intensified. We were inspired to build something that could sense stress early and support the child before the situation becomes overwhelming. CalmBand was born from the idea that technology should quietly support emotional regulation, not react after distress occurs.

What it does CalmBand is an AI-powered, sensor-based wristband designed to help neurodivergent children regulate stress. It continuously monitors physiological and behavioral signals such as heart rate variability, motion patterns, and skin conductance to detect early signs of anxiety or sensory overload. When elevated stress is detected, the band automatically delivers personalized calming feedback—such as gentle vibrations or soothing sounds—while also sending a real-time alert to caregivers through a mobile app.

How we built it The system is designed around a wearable hardware core integrated with intelligent software. The wristband includes sensors for heart rate, motion, and skin response, along with a vibration motor and audio output for feedback. Sensor data is processed using a lightweight machine learning model that learns the child’s baseline patterns and identifies deviations associated with stress. A companion mobile app displays alerts, trends, and allows caregivers to customize calming responses. For the MVP, sensor data and stress detection are simulated to demonstrate the full end-to-end workflow.

Challenges we ran into One major challenge was designing the system in a medically responsible way—ensuring that CalmBand does not claim to diagnose conditions, but instead focuses on detecting stress indicators. Another challenge was balancing continuous sensing with child comfort and simplicity, as the device must be non-intrusive and easy to wear. We also had to carefully frame the AI logic to adapt to individual differences rather than relying on one-size-fits-all thresholds.

Accomplishments that we're proud of We successfully designed a complete end-to-end concept that integrates hardware sensing, AI-driven analysis, real-time caregiver alerts, and on-device calming responses. Despite time constraints, we built a clear technical architecture and MVP flow that demonstrates how CalmBand can work in real-world scenarios. Most importantly, we created a solution that prioritizes empathy, safety, and ethical use of technology.

What we learned Through this project, we learned how critical personalization is when building assistive technology. We also gained insights into designing AI systems that support humans rather than replace them, especially in sensitive contexts involving children. Additionally, we learned how to translate a social problem into a viable robotics-based startup concept with both technical depth and real-world impact.

What's next for CalmBand Next, we plan to test CalmBand with real sensor data and refine the stress-detection model using supervised learning. We also aim to collaborate with therapists and special educators to improve feedback mechanisms and validate effectiveness. Future versions will focus on improving battery life, comfort, and expanding deployment in therapy centers and special education schools.

Built With

• ai
• assistivetechnology
• healthcare
• human-centereddesign
• iot
• ml
• robotics
• wearables