CARES (Cognitive Assistance and Response Emergency System) is an assistive prototype designed to support visually impaired individuals during emergencies using ESP32, MAX30102, NEO-6M GPS, Python backend, SQLite and a web dashboard. It analyses heart-rate variations against personal baselines using deviation, trends, persistence and recovery to classify LOW, MEDIUM or HIGH risk and trigger guardian SMS alerts. Preliminary software testing and simulated evaluation demonstrate its functionality; however, hardware readings remain intermittent, continuous monitoring is not established, and clinical validation is pending.
Introduction
The text presents CARES (Cognitive Assistance and Response Emergency System), a wearable-based assistive technology designed to support visually impaired individuals by monitoring their heart rate, tracking their location, and providing timely alerts to guardians during potentially abnormal situations.
Visually impaired individuals often depend on assistive technologies for navigation and mobility. However, these technologies may not provide sufficient support when a person feels unwell or requires emergency assistance. To address this problem, CARES integrates wearable sensors, an ESP32 microcontroller, GPS technology, and a web-based monitoring platform to provide physiological monitoring and remote assistance.
The proposed CARES system consists of three major hardware components:
ESP32 DevKit V1: Acts as the main controller, processes sensor information, and transmits data through Wi-Fi.
MAX30102 Sensor: Measures heart-rate observations using optical photoplethysmography (PPG) technology.
NEO-6M GPS Module: Provides geographical location coordinates to help guardians identify the user's location.
The system works by collecting heart-rate readings and GPS coordinates and transmitting them to a Python-based backend through HTTP requests. The backend validates the incoming data, stores information in an SQLite database, and processes valid physiological observations using a personalized risk-assessment engine. The results are displayed on a browser-based dashboard.
The proposed system includes the following major functionalities:
Continuous Physiological Monitoring: Collects heart-rate observations using the MAX30102 sensor.
Location Tracking: Uses GPS coordinates to identify the user's geographical location.
Personal Baseline Calculation: Establishes an individual heart-rate reference using trusted observations rather than relying on fixed population-based thresholds.
Risk Assessment: Evaluates heart-rate deviations, percentage changes, and rate of change to identify potentially abnormal conditions.
Explainable Risk Classification: Categorizes the user's condition into three levels:
LOW – No significant abnormality detected.
MEDIUM – Moderate deviation requiring attention.
HIGH – Significant or persistent deviation requiring guardian notification.
Guardian Notification: Initiates configured alerts through SMS when predefined risk conditions are satisfied.
An important feature of CARES is its personalized risk-assessment mechanism. Initially, the system requires at least 300 seconds of trusted observations to establish the user's heart-rate baseline. It then evaluates incoming readings using deviation, percentage deviation, and rate-of-change parameters.
The system also uses persistence checks and confidence assessment to avoid generating alerts because of isolated or temporary sensor fluctuations. A risk level is increased only when abnormal readings persist for a specified number of consecutive samples. This improves the reliability and transparency of the alert mechanism.
The literature review examines existing research on wearable physiological monitoring, stress detection, IoT-based health monitoring, AI-powered assistive technologies, and navigation systems for visually impaired individuals. It identifies a research gap in combining personalized physiological monitoring, GPS-based location tracking, explainable risk assessment, and staged guardian notifications within a single assistive system.
The proposed system is implemented using the following technologies:
Hardware: ESP32, MAX30102, and NEO-6M GPS.
Programming: Python for backend development.
Database: SQLite for storing user information and monitoring events.
Communication: Wi-Fi and HTTP requests for data transmission.
Notification: TextBee SMS gateway for guardian alerts.
Web Interface: Browser-based dashboard for monitoring and visualization.
The preliminary evaluation demonstrates the major system functionalities, including user authentication, hardware interfacing, GPS location display, risk classification, and SMS notification during simulated high-risk conditions. The notification system also distinguishes between generated requests, provider-accepted messages, and confirmed delivery to improve communication transparency.
Conclusion
CARES presents a prototype framework that connects wearable physiological sensing and GPS-based location information with embedded communication, backend processing, explainable risk assessment and guardian-related actions for visually impaired individuals. The ESP32 integrates the MAX30102 and NEO-6M modules and transmits available observations to a Python backend, where validated heart-rate samples are assessed relative to a personal baseline. The decision engine considers deviation, trend, persistence, recovery and evidence confidence before producing LOW, MEDIUM or HIGH states. These states are displayed through a browser dashboard and mapped to monitoring or notification actions. The software evaluation included 173 automated tests, with 172 passing in the reported snapshot, and a simulated-cohort comparison in which CARES reduced state oscillations while introducing trade-offs in latency and false-positive rate. A synthetic demonstration exercised the guardian workflow and produced one SMS on a team-controlled phone. Nevertheless, sensor readings were intermittent, continuous monitoring was not established, and the prototype has not been clinically validated. CARES should therefore be regarded as an exploratory assistive system rather than a diagnostic or emergency service.
References
[1] S. Cano, C. Cubillos, R. Alfaro, A. Romo, M. García, and F. Moreira, “Wearable Solutions Using Physiological Signals for Stress Monitoring on Individuals with Autism Spectrum Disorder (ASD): A Systematic Literature Review,” Sensors, vol. 24, no. 24, Art. no. 8137, 2024, doi: 10.3390/s24248137.
[2] A. Pinge, V. Gad, D. Jaisighani, S. Ghosh, and S. Sen, “Detection and Monitoring of Stress Using Wearables: A Systematic Review,” Frontiers in Computer Science, vol. 6, Art. no. 1478851, 2024, doi: 10.3389/fcomp.2024.1478851.
[3] M. S. A. Baig et al., “AI-Based Wearable Vision Assistance System for the Visually Impaired,” arXiv preprint arXiv:2412.20059, 2024. Available: https://arxiv.org/abs/2412.20059.
[4] G. Vos et al., “Generalizable Machine Learning for Stress Monitoring from Wearable Devices: A Systematic Literature Review,” arXiv preprint arXiv:2209.15137, 2022. Available: https://arxiv.org/abs/2209.15137.
[5] M. Gjoreski, H. Gjoreski, M. Luštrek, and M. Gams, “Continuous Stress Detection Using a Wrist Device: In Laboratory and Real Life,” in Proceedings of the 2016 ACM International Joint Conference on Pervasive and Ubiquitous Computing: Adjunct, 2016, pp. 1185–1193, doi: 10.1145/2968219.2968306. DBLP+1
[6] S. Nath et al., “IoT-Based Smart Health Monitoring System Using Wearable Sensors,” IEEE Journal of Biomedical and Health Informatics, 2024.
[7] R. Karthik et al., “Real-Time Health Monitoring and Alert System Using ESP32,” International Journal of Engineering Research and Technology (IJERT), vol. 14, no. 2, 2025.
[8] H. Lee et al., “AI-Driven Wearable Cognitive Monitoring and Emergency Response System,” IEEE Access, vol. 13, 2025.
[9] S. Lokesh, P. Sukesh, S. Tejaswini, M. Meera, and G. Udhayakumar, “IoT-Powered Wearable Assist Device for Visually Impaired,” Disability and Rehabilitation: Assistive Technology, vol. 21, no. 4, pp. 1409–1420, 2026, doi: 10.1080/17483107.2025.2582033.
[10] G. I. Okolo, T. Althobaiti, and N. Ramzan, “Assistive Systems for Visually Impaired Persons: Challenges and Opportunities for Navigation Assistance,” Sensors, vol. 24, no. 11, Art. no. 3572, 2024, doi: 10.3390/s24113572.
[11] J. S. Sunny, C. P. K. Patro, K. Karnani, S. C. Pingle, F. Lin, M. Anekoji, L. D. Jones, S. Kesari, and S. Ashili, “Anomaly Detection Framework for Wearables Data: A Perspective Review on Data Concepts, Data Analysis Algorithms and Prospects,” Sensors, vol. 22, no. 3, Art. no. 756, 2022, doi: 10.3390/s22030756.
[12] P. Arpaia, F. Crauso, E. De Benedetto, L. Duraccio, G. Improta, and F. Serino, “Soft Transducer for Patient’s Vitals Telemonitoring with Deep Learning-Based Personalized Anomaly Detection,” Sensors, vol. 22, no. 2, Art. no. 536, 2022, doi: 10.3390/s22020536.