Join the 155,000+ IMP followers

www.usa-automation.net

NEXCOM Showcases Safe Embodied AI Robots at NVIDIA GTC 2026

AI robot controllers on NVIDIA IGX and Jetson platforms integrate functional safety, real-time motion control and ISA/IEC 62443 cybersecurity.

  www.nexcom.com
NEXCOM Showcases Safe Embodied AI Robots at NVIDIA GTC 2026
Reference Image

In the industrial robotics and AI-powered robot control sectors, NEXCOM and its subsidiary NEXCOM Robotic Solutions (NexCOBOT Co. Ltd.) will present safety-focused embodied AI solutions at NVIDIA GTC 2026, held March 16–19, 2026 at the San Jose McEnery Convention Center, San Jose, CA, USA (Booth 166).

The showcase centers on the integration of AI inference, deterministic motion control and certified functional safety within a unified robotic controller architecture.

Safety-Enhanced Robot Controller on NVIDIA IGX
The RCB400 T20-05 all-in-one controller, powered by the NVIDIA IGX T5000 module, demonstrates how AI-based perception and multi-axis safety control can operate within a real-time robotic system.

The architecture combines:
  • NexRTOS for deterministic real-time robot motion
  • NexECM for EtherCAT motion control
  • EBK 210 Safety MCU expansion board
  • Synapticon POSITRON Safety AI software
The system fuses data from a RealSense 3D camera and safety LiDAR to detect dynamic obstacles in real time. Functional safety features native to the IGX platform are activated via the safety MCU, enabling preventive safe motion control to reduce human-robot collision risks.

Multi-master EtherCAT communication ensures coordinated control of safety-certified joints, supporting collaborative robotics scenarios where robots operate in proximity to humans.


NEXCOM Showcases Safe Embodied AI Robots at NVIDIA GTC 2026

AI Simulation and Digital Twin Integration
A second solution integrates virtual simulation and physical deployment using NVIDIA Isaac Sim.

The setup includes:
  • MARS400 T10-05 robotic controller powered by NVIDIA Jetson T5000
  • AS-531AW-TC server for visual data processing
  • High-fidelity digital twin robot model in Isaac Sim
This configuration demonstrates AI-based bin-picking within a simulated environment, enabling robot modeling, motion optimization and system validation before physical commissioning.

By validating control logic and AI inference in a digital twin environment, developers can reduce development time and deployment costs while improving system precision.

Embedded Cybersecurity for Robotic Controllers
To address increasing cybersecurity risks in connected robotics, NEXCOM presents eSAF Guardian, an embedded cybersecurity agent for RTOS and Linux-based robotic systems.

The solution monitors:
  • Running processes
  • File access
  • Network connections
Operating at a low system level, eSAF Guardian detects and blocks unauthorized activities, mitigating malware and intrusion risks. The design supports compliance with ISA/IEC 62443 industrial cybersecurity standards.

Integrated Approach to Functional Safety and AI
The presented solutions address a key challenge in embodied AI robotics: balancing high-performance AI processing with deterministic motion control and certified functional safety.

By integrating AI inference, EtherCAT-based motion control, digital twin simulation and embedded cybersecurity into cohesive controller platforms, NEXCOM targets industrial and humanoid robot developers requiring scalable, safety-oriented architectures.
 

  Ask For More Information…

LinkedIn
Pinterest

Join the 155,000+ IMP followers