NEXCOM Group has introduced a humanoid robot controller designed to integrate artificial intelligence computing, motion control, and functional safety into a single platform. The company unveiled the MARS400 T10 controller at the Taiwan Automation Intelligence and Robot Show this week.
The controller is built on NVIDIA’s Jetson Thor platform and incorporates NEXCOM’s in-house functional safety system, ESC210. According to the company, the modular design reduces the latency and space requirements associated with traditional multi-platform robot architectures, while enabling improved movement precision, extended operating times, and more compact robot designs.
NEXCOM stated that current robot systems generally require separate processors for AI computing, motion control, and safety functions, which complicates development and can result in delays during operation. By consolidating these components, the MARS400 T10 aims to shorten development cycles and improve real-time performance, particularly for humanoid robots.
The platform supports computing performance of up to 2070 FP4 TFLOPS and is compatible with NVIDIA’s CUDA-accelerated robotics libraries and AI models. The company said this integration allows manufacturers to more quickly bring safety-certified robots to market.
In 2024, NEXCOM received TÜV Rheinland functional safety certification, which it says makes it the only company in Taiwan to provide modularized robotic safety solutions compliant with EN 61508 and EN ISO 13849-1 standards. The certification covers a suite of safety components, including masters, modules, and I/O interfaces. In June 2025, NEXCOM was invited to join NVIDIA’s Halos AI Systems Inspection Lab to contribute to industry safety advancements.
At the Taiwan Automation Intelligence and Robot Show, NEXCOM is presenting under the theme “NEXCOM Empowerment, Driving AI Smart Manufacturing Leap Forward.” Its exhibits highlight applications of AI across several areas, including humanoid robotics, motion control, industrial IoT monitoring, augmented reality for mobile robots, digital twins, enterprise systems, and healthcare solutions.
