ROHM Semiconductor has announced the development of AI-equipped microcontrollers designated ML63Q253x-NNNxx and ML63Q255x-NNNxx that perform fault prediction and degradation forecasting using sensor data in industrial equipment including motors. These MCUs execute both learning and inference operations without requiring network connectivity, operating as standalone devices for real-time monitoring applications.
The MCUs implement ROHM’s Solist-AI technology using a 3-layer neural network algorithm that enables on-device learning and inference. Unlike cloud-based AI systems that require network connectivity for training and inference, or edge AI implementations that rely on cloud-device combinations, these MCUs perform both functions locally. This approach eliminates network latency issues and reduces security risks associated with cloud-dependent systems.
Each MCU integrates a 32-bit Arm Cortex-M0+ core operating at a maximum frequency of 48 MHz alongside ROHM’s proprietary AI accelerator called AxlCORE-ODL. The AI accelerator delivers processing speeds approximately 1,000 times faster than software-based implementations when operating at 12MHz, enabling real-time anomaly detection with numerical output capabilities. The on-device learning functionality allows the system to adapt to installation environment variations and unit-to-unit differences within the same equipment model.
The hardware configuration includes a CAN FD controller, 3-phase motor control PWM functionality, and dual 12-bit analog-to-digital converters. Power consumption measures approximately 40 mW during operation. The MCUs support various control and data processing applications across industrial equipment, residential facilities, and home appliances through multiple serial interfaces and versatile timer functions.
The product lineup encompasses 16 variants with different memory configurations, package types, pin counts, and packaging specifications. Mass production of eight models in TQFP packages began in February 2025. Two models featuring 256KB Code Flash memory with taping packaging are available through online distributors along with MCU evaluation boards for development purposes.
ROHM provides development support through an AI simulation tool called Solist-AI Sim that enables users to evaluate learning and inference effectiveness before deployment. The simulation tool generates data that serves as training input for the actual MCU implementation. The development environment maintains compatibility with standard Arm core tools and includes ROHM’s integrated development environment, AI operation verification simulator, and real-time effectiveness assessment viewer.
