Heatwave Guardian
Heatwave Guardian is a working edge-AI thermal risk monitoring prototype built for Arduino hardware and Edge Impulse.
NNARASIMA MAHARANA- Problem, approach and result explained
- Expert reviewed· 6 reviews
90%
Alert & precaution System
86%
ML Model testing output(Edge Impulse)
70%
Enclosure Designing
Overview
The problem
Elderly individuals living alone face severe, silent risks during extreme winter nights. Prolonged exposure to low indoor temperatures can quickly lead to dangerous cold stress, yet older adults often have reduced thermal sensitivity and may not realize their home has dropped to unsafe levels. Worse, if an elderly person is asleep, has fallen, or becomes incapacitated during a sudden nighttime temperature drop, they cannot reach a phone or press a manual SOS button, leaving no one nearby to notice something is wrong.
The approach
We built this project which is an affordable, zero-interaction safety-support system designed to protect elderly residents living alone. Instead of relying on simple fixed-threshold alarms that cause frequent false positives, we set out to use Edge AI on the Arduino UNO Q to combine continuous indoor temperature, environmental monitoring, humidity data, wet bulb temperature to collect data and automatically recognizing prolonged cold exposure and alerting a caregiver even when the resident cannot respond.
The result
Yes our working proof-of-concept successfully runs end-to-end on-device Edge AI inference, triggers local audio/visual warnings, with actuation of protective system(heater) and pushes real-time caregiver alerts message to both the monitoring dashboard and the caregiver without requiring any manual input from the resident.
Reflection
Currently, we have not yet integrated the movement sensor or fall detection module into the prototype. Additionally, while the system collects indoor humidity data, it is presently used only to calculate wet-bulb temperature for more accurate temperature assessment rather than actuating a humidifier or humidity-control device.
Submitted to Heat Intelligence at the Edge
- Arduino board used?
Yes
- Edge AI used?
Yes
- AI/ML model used
A compact, fully connected Neural Network classifier developed using Edge Impulse, intended to classify environmental conditions into Normal, Moderate Heat Risk, and High Heat Risk based on temperature and humidity sensor data.
- Sensors used
SHT41: Measures indoor temperature and relative humidity. DS18B20 waterproof temperature sensor: Measures outdoor temperature, WBGT . DHT11: Provides an additional temperature and humidity reading for monitoring and comparison in side the indoor room.
- Camera used
None
- Modulino used
No Modulino module is currently used in this project. The sensors and output devices are connected directly to the Arduino UNO Q through the required GPIO and I²C interfaces.
- Problem being solved
High temperatures can make indoor environments uncomfortable and potentially unsafe, especially when people are not actively monitoring the conditions. Heatwave Guardian addresses this problem by continuously monitoring environmental conditions, automatically controlling a cooling fan, and providing an alert when the outdoor temperature exceeds the configured threshold of 50°C.
- AI model training platform
Edge Impulse Studio, using a labelled dataset collected from the SHT41, DHT11, and DS18B20 sensors.
- Arduino UNO Q features used
The project uses the Arduino UNO Q to interface with multiple sensors, process temperature and humidity readings, control the DC fan through an L298N motor driver, activate the buzzer, and update the OLED display. Python-based application logic and the Arduino Router Bridge communication framework are used to exchange sensor readings and control commands with the dashboard.
- Deployment method
The system is deployed as a hardware prototype using the Arduino UNO Q, environmental sensors, an OLED display, a buzzer, and a DC fan. The Arduino sketch handles sensor acquisition and hardware control, while the Python application provides monitoring and control logic through the dashboard. A battery pack and solar panel are included in the proposed design for solar-assisted operation, subject to proper power-system integration and testing.
- Key innovation
The system is deployed as a hardware prototype using the Arduino UNO Q, environmental sensors, an OLED display, a buzzer, and a DC fan. The Arduino sketch handles sensor acquisition and hardware control, while the Python application provides monitoring and control logic through the dashboard. A battery pack and solar panel are included in the proposed design for solar-assisted operation, subject to proper power-system integration and testing.
- Solution summary
Heatwave Guardian is an Arduino UNO Q-based environmental monitoring and cooling-control system designed to respond to high-temperature conditions. It measures temperature and humidity, displays readings on an OLED screen, monitors data through a software dashboard, controls a DC fan automatically, and activates a buzzer when the outdoor temperature exceeds 50°C. The system aims to make heat monitoring more convenient and reduce the need for continuous manual supervision.
- Report (PDF)
- Report (PDF)PDF
- What's working now?
Coldwave Guardian is a working edge-AI cold risk monitoring prototype built for Arduino Q hardware and Edge Impulse. Current working mechanism are On-Device Edge Inference, Live Monitoring Dashboard, instantaneous forecasting Indications, Prediction actuation system, Alerts message by GSM.
- Deck or report (PDF)
- Deck or report (PDF)PDF
AI tools used
Real time data refinement and labelling