STMicroelectronics introduces control chips for robots and a unified STM32 platform
The first three STM32A chips are sampling to customers. ST’s pitch combines reusable software with control hardware designed for safety-critical machines.
Loading page…
The first three STM32A chips are sampling to customers. ST’s pitch combines reusable software with control hardware designed for safety-critical machines.
Listen to this story
The three Cortex-M-based STM32A variants divide control work by complexity: A4 targets automotive local nodes and safety systems, A5 mid-range safety-critical applications, and A9 multi-domain control. Customer sampling is underway, but broader availability is planned for 2027. ST’s unified platform is intended to help developers reuse architecture and software across industrial, automotive, and physical-AI machines. For robotics, ST and Arm emphasize predictable, redundant embedded control near sensors and actuators, designed to keep systems operating safely when components fail.
STM32A4 targets automotive local control nodes and safety systems; STM32A5 targets mid-range safety-critical applications; STM32A9 handles safety-critical control across multiple domains.
The first three chips are being sampled to customers, while broader availability is planned for 2027.
Arm divides a robot’s computing needs into perception, interaction, and actuation, with embedded processors handling fast, predictable control near sensing and movement.
STMicroelectronics is bringing industrial, automotive and robotics control into a unified STM32 microcontroller platform. Announced November 4 at the STM32 Summit, the strategy includes the new Cortex-M-based STM32A chip series. The first three chips are sampling to customers, while broader availability is planned for 2027—a staged rollout for hardware aimed at controlling safety-critical machines.
ST says the platform will provide a consistent architecture and software foundation, allowing developers to reuse work across applications. Its convergence strategy spans industrial equipment, vehicles and physical AI: systems that sense their surroundings, decide what to do and act in the real world.
Arm, whose Cortex-M processor architecture underpins STM32A, explained the engineering problem in a November 5 article about the collaboration. A robot picking up a glass must recognize it, locate it, plan a reach and calculate enough force to hold it without breaking it. Sensors keep checking position, pressure and movement as the control system adjusts.
That sequence involves different kinds of computing. Arm separates them into perception, interaction and actuation—the execution of physical movements. Perception handles understanding the surroundings; interaction handles instructions such as voice and gestures. Actuation controls motors, joints and other moving components. Cloud or data-center computing supports fleet management, simulation and training.
The control layer has a particular demand: fast, predictable responses as conditions change. Arm places embedded processors at this point, close to sensing and movement. It says AI and conventional software must coordinate at different speeds, while the whole machine stays within limits on power, weight, heat and cost.
ST’s initial STM32A lineup addresses several levels of control rather than a single robot design. The company assigns the first three products these roles:
ST says the first three STM32A microcontrollers are being sampled to customers. Broader availability remains a planned milestone for 2027, not a completed rollout.
For Arm, the collaboration also reaches beyond these initial products. It says its work with ST is helping shape the next generation of the Cortex-M roadmap around growing demands for sensing, processing and control. The stated aim is to put more intelligence near the point where machines interact with the world, while retaining efficiency, reliability and responsiveness.
Loading discussion...
Join the conversation
Explain where shared engineering helps—and where you would want separate approaches.
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.