Renesas Electronics Reality AI

Overview

Reality AI is a development environment for developing edge AI software that combines signal processing, machine learning, and anomaly detection that can be run on Renesas MCU/MPU cores.
By simply uploading training data, whose feasibility is difficult to judge, to the cloud, it is possible to display the achievable accuracy, retrain on the cloud, and perform grid searches of sensor positions.

Source: https://www.renesas.com/jp/ja/blogs/year-after-acquisition-renesas-redefining-ai-solutions-portfolio-reality-ai-software

Features

・Signal processing technology
Reality AI can compare and validate algorithms across a wide range of frequency and time domain feature spaces and automatically select the optimal algorithm.

・AI technology
Reality AI can automatically create machine learning models. The generated machine learning models prevent black Box by visualizing the learning results, and at the same time provide various analysis tools for hardware optimization for production.

・Machine learning models for embedded systems
The Reality AI tool can generate trained machine learning models for each MCU/MPU, such as Renesas RX, RL78, RA, and RZ, and then link them directly to existing projects to implement functions that use machine learning.