CardioAlert

Inspiration

A close friend of mine lost both his nana and mama due to insufficient medical emergency care. This heart-wrenching loss highlighted the severe need for a system that not only senses threats to health but also reacts intelligently and promptly by sending out emergency notifications. Inspired by this tragedy, we created CardioAlert—a wearable device that continuously monitors heart rate in real time and uses an AI-driven engine to analyze data patterns, ensuring that guardians and emergency services are alerted immediately during critical cardiac events. We are committed to building an innovative system that harnesses technology and artificial intelligence to save lives through timely intervention.

What It Does

CardioAlert is a real-time heart rate monitoring system designed to detect potentially life-threatening changes in heart rate. Utilizing a pulse rate sensor, it continuously tracks the user's heart rate and, when readings exceed a predefined threshold, the system employs its AI module to cross-verify these signals with historical patterns and additional biometric data. Once the AI confirms a genuine risk, CardioAlert instantly sends an SMS alert to the appointed guardian along with the user’s current GPS location. This intelligent early alerting system not only speeds up emergency response but also minimizes false alarms, ensuring help reaches the user as quickly as possible.

How We Built It

CardioAlert is built around an ESP32 microcontroller that manages both processing and connectivity, enhanced by a SIM module to dispatch SMS alerts. A GPS module provides precise location tracking, while the HW827 pulse rate sensor continuously captures the heart rate. What sets our project apart is the integration of an AI engine within the app. This AI system uses models such as LSTM networks for time-series analysis and unsupervised learning techniques for anomaly detection, which cross-check sensor data against established personal patterns. This multi-layered approach ensures that only verified, critical alerts trigger emergency responses, combining hardware reliability with smart software analytics.

Challenges We Faced

One of our biggest hurdles was achieving accurate calibration of the HW827 pulse rate sensor. In the early stages, inconsistent sensor readings and unfamiliar libraries for interfacing with the ESP32 posed significant challenges. Integrating the AI components added another layer of complexity—ensuring that the machine learning models were properly trained and validated to distinguish between normal fluctuations and genuine emergencies. Despite these obstacles, we persevered, fine-tuning both our hardware calibration and AI algorithms to achieve reliable, real-time monitoring.

Achievements We Are Proud Of

We successfully completed the project within the specified timeframe, overcoming technical and calibration issues to deliver accurate heart rate measurements. Our prototype now seamlessly integrates sensor data with AI-driven analysis to provide real-time, validated alerts. As a result, guardians receive prompt SMS notifications complete with GPS data whenever a potential cardiac emergency is detected. This breakthrough not only demonstrates our technical proficiency but also reinforces our commitment to saving lives through technology.

What We Learned

Throughout the development of CardioAlert, we gained invaluable insights into sensor calibration, real-time data processing, and the integration of AI with embedded systems. We learned how to troubleshoot and optimize both hardware and software components, and we discovered the immense potential of data-driven decision-making in healthcare. Most importantly, this project taught us that combining traditional engineering with advanced AI can significantly enhance emergency response, ultimately making a real impact on people’s lives.

What's Next for CardioAlert

In our next phase, we aim to further improve the accuracy of heart rate monitoring and extend the system's capabilities to track additional biometric parameters. We plan to enhance our AI algorithms by incorporating more advanced models and continuous learning techniques, such as federated learning, to adapt to individual user patterns over time. Future updates will also focus on direct integration with hospital systems and emergency services, ensuring that medical professionals can respond even faster. Additionally, we are exploring the possibility of detecting other medical emergencies through supplementary sensors, further broadening CardioAlert’s life-saving potential.