Introduction of image recognition reference ECU (JAOiECU)

As a solution to the social issues of a declining birthrate and aging population, as well as a lack of successors, there are high hopes for efficiency and labor savings through automation support.
Our image recognition reference ECU incorporates the functions required for AI in a durable, waterproof and dustproof housing, enabling image recognition to be used in all types of mobility.

It uses an industrial module that extends the functionality of the NVIDIA® Jetson AGX™ Orin system-on-module.
The Jetson AGX Orin Industrial module delivers AI performance of 248 trillion operations per second (248 TOPs: INT8) in a power-efficient design.

It uses a Gigabit Multimedia Serial Link SerDesIC (GMSL2) that is used in automotive applications.
By incorporating a dedicated 6-channel GMSL camera interface without using USB or Ethernet, it solves two trending demands: "higher image quality" and "increased number of cameras."

It uses a robust housing that is waterproof and dustproof (IP69K) for outdoor environments.
From farm fields (rice paddies and fields) to construction sites, manufacturing sites and factories, it protects equipment and ensures stable operation even in harsh environments where not only vibration and shock but also water, oil, dust and dirt are flying around.

By using the NVIDIA JetPack SDK, you can achieve the same usability as your usual development environment and use your existing software assets as they are.

You can test the reference ECU, including the camera, in advance and confirm its operation in the actual environment before purchasing or developing a custom reference ECU.

• Image transmission interface GMSL (Analog Devices)