Industrial Robot Safety Control Architecture

NEXCOM Robotic Solutions participated in NVIDIA GTC Berlin, held on October 21–22, 2026, at STATION-Berlin in Germany, to present safety control frameworks and open architecture platforms for next-generation robotics.

Core Technology and System Architecture
The demonstrated architecture centers on the RCB400 humanoid robot controller, which integrates the NVIDIA IGX Thor edge AI platform to process high-throughput perceptual workloads and real-time motion control commands simultaneously. The system runs on the proprietary NexRTOS real-time operating system combined with NexECM EtherCAT motion control software to govern multi-joint synchronization. Spatial data acquisition relies on a RealSense D555 depth camera that tracks human positioning parameters within the workspace. To establish hardware-level redundancy, an EBK Safety MCU expansion board monitors the state of the central controller. If an anomaly occurs, the safety microcontroller intervenes independently to execute native functional safety routines of the NVIDIA IGX Thor platform, complying with IEC 61508 and ISO 13849 requirements.

Deployment and Implementation
The solution is implemented as a pre-integrated modular framework through the RoboWIZ AI Robot Open Platform. Deployment involves connecting industrial-grade vision sensors, such as RealSense and StereoLabs units, alongside Synapticon ACTILINK-JD08 high-dynamic joint modules that communicate via EtherCAT protocols. System integrators deploy the hardware architecture directly into humanoid robot chassis, while functional safety engineering frameworks provided by partners like Reynolds & Moore manage hazard assessment and safety validation workflows, reducing custom integration cycles.

Industry Applications
Target sectors include smart manufacturing, automated logistics, and commercial service environments where autonomous systems operate adjacent to personnel. Specific use cases include dynamic speed reduction based on proximity calculations and hard-stop intervention when operators cross defined hazardous boundary thresholds. These mechanisms maintain process safety while preventing mechanical overload or uncontrolled trajectory deviations.

Operational Impact and Performance Metrics
Integration of edge AI processing and hardware-level safety monitoring reduces latency in human-detection feedback loops. The dual-core architecture separates standard OS functions from deterministic safety routines, ensuring that safety-rated interrupts execute without bus congestion. The inclusion of hardware safety microcontrollers and pre-validated open-platform components minimizes custom verification steps, lowering compliance timelines for international machinery directives.

Edited by Sucithra Mani, Induportals editor – adapted by AI

www.nexcom.com