This device predicts when a refrigerator may fail using built-in ML

Arduino Team — July 4, 2022

The refrigerator is one of the centerpieces of a modern kitchen, and loss of cooling can result in hundreds or even thousands of dollars in spoiled produce. Perhaps most importantly, a sudden loss of medicines or vaccines that rely heavily on refrigeration can have a big impact on the people who need them. Swapnil Verma wanted to solve this problem. So he came up with the idea of ​​integrating a simple machine learning model into a device that could monitor failures.

When collecting data points for model training, Verma started by identifying different failure modes, such as a drop in temperature, a change in humidity, or just an anomaly. He chose to use an Arduino Nano 33 BLE Sense with its built-in temperature/humidity and ambient light sensors. From there, the data is streamed via Bluetooth® LE to a Portenta H7 and saved to a microSD card. Verma then uploaded the resulting CSV files to Edge Impulse Studio and trained an anomaly detection model that could recognize when conditions inside the refrigerator are incorrect.

Although the deployment does not currently involve sending alerts, Verma suggested that the functionality could be added in the future, especially for the medical field. Want to dive into the details of the project? Check out his Edge Impulse tutorial as well as