Low-Cost Flexible AI Platform With Power Of Raspberry Pi and ESP32

- Advertisement -

It also features a 9-axis IMU and digital microphone that can be effectively implemented for a range of IoT applications

Combining the power of Raspberry Pi and ESP32 to help users develop compelling Edge AI projects is the new UDOO KEY board by UDOO, a company focused on easing human lives with technology. 

- Advertisement -

Designed for professionals and hobbyists alike, the UDOO KEY is 100% compatible with an accessory available for Raspberry Pi Pico and with OLIMEX UEXT modules for ESP32.

This flexible AI platform is fully programmable that’s built upon a Raspberry Pi RP2040 dual ARM Cortex-M0, with a QSPI 8MB Flash, a 133MHz clock and 264KB of on-chip SRAM. 

The ESP32, on the other hand, is based on a dual-core Xtensa 32-bit LX6, with a 16MB flash memory, an 8MB PSRAM, Wi-Fi and Bluetooth. 

This amalgamation of best of the both worlds enables machine learning applications in the most popular programming languages and libraries, including TinyML, TensorFlow Lite, MicroPython, C, C++, etc. 

Other features include:

• 9-axis IMU + digital microphone for advanced sensing (only available with UDOO KEY PRO)
• SWD interface communication
• Internal Temperature sensor – RP2040
• 26 GPIOs, 8x PIO with I2C, UART, PWM and SPI interfaces – RP2040
• UEXT connector(I2C, SPI, UART) – ESP32
• Onboard battery charger
• 130mm x 40mm size

You can program the board via USB-C and decide whether to talk with Raspberry Pi or ESP32 via jumper. But because ESP32 does not naturally mount a USB port, a UART-to-USB bridge is present. 

Software Support

The UDOO KEY comes with built-in tools to make development as smooth as possible. The microcontroller comes with dedicated, fully open-source firmware based on Free RTOS and with ready to use SDKs. 

UDOO KEY natively supports Clea, which is an AI & IoT platform that allows data collection and management from connected devices for creating flows tailored to the user’s needs. It also provides visualisation tools to turn IoT data into actionable insights. 

Clea is built on top of two open-source projects: Astarte and Edgehog. While Astarte allows the developers to manage and process IoT data, Edgehog enables fleet device management and OTA updates, enhancing ESP32’s firmware with a full-fledged device lifecycle solution. 

The AI-based projects, for both industrial and non-industrial scenarios that you can build with UDOO KEY are endless. Some of them include:

• Predictive Maintenance
• Sound Recognition
• Keyword Recognition
• Object Detection & Classification
• Gesture Recognition
• Face Recognition

The UDOO KEY is now available on Kickstarter. It also comes with a one-year subscription to CleaCloud, an Open Access AI platform capable of deploying AI models and applications over a fleet of IoT devices through over-the-air updates.