Welcome to the AI-on-the-edge-device

2023-11-14

Artificial intelligence based systems have been established in our everyday lives. Most systems rely on powerful processors or a direct connection to the cloud. This project brings edge computing into practice by implementing an AI network on an ESP32 device. It allows you to digitalize analog water, gas, power, and other meters using cheap and easily available hardware. The key features of this project include Tensorflow Lite integration, inline image processing, small and cheap device size, integrated camera and illumination, web surface for administration and control, OTA-interface for updates, full integration into Homeassistant, support for Influx DB 1, MQTT, and REST API. The workflow involves taking a photo of the meter at a defined interval, extracting regions of interest, and running them through an AI to obtain the digitalized value of the meter. The digitalized value can be sent to a MQTT broker, written to an InfluxDb, or provided through a REST API. For setup instructions and further information, refer to the documentation and articles mentioned in the text. The latest available version can be downloaded from the Releases page. Flashing of the ESP32 can be done through USB or over-the-air. The SD-Card must be flashed separately. A 3D-printable housing is also available for this project.

Link [ https://github.com/jomjol/AI-on-the-edge-device ]

Previous Article

Welcome to the AI-on-the-edge-device

2023-11-14

Artificial intelligence based systems have been established in our everyday lives. Most systems rely on powerful processors or a direct connection to the cloud. This project brings edge computing into practice by implementing an AI network on an ESP32 device. It allows you to digitalize analog water, gas, power, and other meters using cheap and easily available hardware. The key features of this project include Tensorflow Lite integration, inline image processing, small and cheap device size, integrated camera and illumination, web surface for administration and control, OTA-interface for updates, full integration into Homeassistant, support for Influx DB 1, MQTT, and REST API. The workflow involves taking a photo of the meter at a defined interval, extracting regions of interest, and running them through an AI to obtain the digitalized value of the meter. The digitalized value can be sent to a MQTT broker, written to an InfluxDb, or provided through a REST API. For setup instructions and further information, refer to the documentation and articles mentioned in the text. The latest available version can be downloaded from the Releases page. Flashing of the ESP32 can be done through USB or over-the-air. The SD-Card must be flashed separately. A 3D-printable housing is also available for this project.

Link [ https://github.com/jomjol/AI-on-the-edge-device ]

Copyright © 2024 All rights reserved

Rss

Atom