MindSync
Smart Stress Tracking for Life Saving Actions
Created on 9th March 2025
•
MindSync
Smart Stress Tracking for Life Saving Actions
The problem MindSync solves
Children’s Safety: Detects stress before potential threats like abduction or bullying escalate.
Elderly Care: Alerts caregivers when distress or health risks arise.
Mental Health & Suicide Prevention: Identifies early stress signals, allowing timely intervention.
Addiction Recovery: Monitors withdrawal symptoms and triggers alerts for medical assistance.
Women’s Safety: Detects panic and distress, automatically triggering emergency responses.
Challenges we ran into
Edge AI Model Deployment Constraints:
The ESP32’s limited computational power and memory posed challenges in directly running our machine learning model. Optimizing the model for edge deployment while maintaining high accuracy required extensive quantization and compression techniques, but latency and precision trade-offs persisted.
Sensor Failures & Signal Noise:
The MAX30102 heart rate sensor exhibited inconsistencies due to poor signal acquisition in certain conditions. Factors such as ambient light interference, improper skin contact, and power fluctuations led to unreliable readings, necessitating advanced filtering algorithms and alternative sensor placement strategies.
HTTP Server Communication Issues:
Establishing a stable data transmission pipeline between the ESP32 and the cloud via an HTTP server was challenging. Packet loss, latency in request handling, and intermittent connectivity issues disrupted real-time data flow. Implementing MQTT as an alternative and optimizing request handling mechanisms improved reliability.
Each of these challenges required hardware debugging, firmware optimization, and protocol refinements to ensure seamless integration of real-time stress detection and AI-driven emergency response.