RCB400 and MARS400 Series: Scalable Edge AI Platforms for Robotics and Industrial Automation

17 April 2026 Knowledge Base

NEXCOBOT RCB400 and MARS400: A Scalable Edge AI Platform for Next-Gen Robotics

NEXCOBOT’s RCB400 and MARS400 Series deliver a scalable Edge AI platform for robotics, autonomous mobile robots, industrial machines, and real-time automation. The portfolio is designed to support both deep OEM integration and fast system-level deployment, giving machine builders a clear path from compact board-level design to ruggedized industrial control systems.

clipboard-image-1776451993.png

Two Product Paths, One Scalable Platform

The RCB400 Series is the board-level solution for customers that need maximum flexibility in mechanical design. It is positioned as an ultra-compact Single Board Computer for embedded robotics, compact machine architectures, and customer-defined enclosures.

The MARS400 Series uses the same hardware core inside a ruggedized industrial enclosure with optimized cooling and ready-to-deploy system integration. This makes it ideal for faster time-to-market in industrial robotics and automation projects where reliability, thermal design, and deployment speed are critical.

Built for Modern AI Robotics

The series is aimed at advanced robotics use cases such as humanoid robots, industrial robot arms, AMRs, and AI-enabled automation systems. NEXCOBOT presents the platform as a way to combine AI computing, real-time control, camera connectivity, industrial I/O, and future-ready expansion on one common architecture.

The hardware roadmap ranges from compact Orin Nano and Orin NX solutions to AGX Orin and Jetson Thor-based high-performance platforms. That range is important for OEMs that want to scale from entry-level embedded control to premium AI robotics without changing to a different product family.

Featured Products

  • RCB400 T10-05
    Board-level AI control platform based on NVIDIA Jetson AGX Thor T5000, developed for top-tier robotics and embodied AI workloads with up to 2,070 TFLOPS class performance in the Thor family.
  • MARS400 T10-05
    System-level modular AI controller for robotics and automation, combining Thor-class computing with an industrial enclosure for faster deployment.
  • RCB400 A10-04-C
    All-in-one AI control board based on NVIDIA Jetson AGX Orin 64GB, positioned for high-performance industrial robotics, machine vision, and control integration at 275 TOPS.
  • MARS400 A10-04-C
    AI robotic control platform built around AGX Orin, offering strong compute density, industrial I/O, and system-level readiness for automation deployments.
  • RCB400 N00-02
    Compact Orin NX 16GB board-level platform rated at 100 TOPS, well suited for mobile robots, embedded AI, and space-constrained mechanical integration.
  • MARS400 N00-02
    System-level Orin NX platform that brings compact AI performance, industrial connectivity, and deployment-ready packaging to edge robotics applications.
  • RCB400 N02-01
    Orin Nano-based board for lower-power intelligent systems, offering an accessible entry point into the same product ecosystem.
  • MARS400 N02-01
    Compact system solution for practical AI edge deployment in cost-sensitive robotics and automation projects.

Performance Range from 20 TOPS to Thor-Class AI

The MARS400 and RCB400 matrix covers a wide performance spectrum, from 20 TOPS entry configurations up to Thor-based platforms positioned at 2,070 TFLOPS in the highest tier shown in the sales material. This gives product teams a common family for low-power embedded systems, mid-range industrial AI, and flagship robotics controllers.

The AGX Orin class is presented at up to 275 TOPS, while Orin NX and Orin Nano models address the 100 TOPS, 70 TOPS, 67 TOPS, 40 TOPS, and 20 TOPS segments. For customers planning long product lifecycles, that lineup supports a clean migration path across multiple performance levels.

clipboard-image-1776451902.png

Key Hardware Advantages

  • Shared product architecture from board-level integration to full industrial system deployment.
  • Unified enclosure across performance tiers to simplify upgrades without mechanical redesign.
  • Professional thermal design and airflow optimization for stable high-performance workloads.
  • Locking connector strategy to improve cable retention in mobile robotics and vibration-heavy environments.
  • Modular expansion through EBK modules for networking, storage, USB, serial, digital I/O, and high-speed connectivity.

Expansion Ecosystem for Custom Systems

A major strength of the platform is the expansion ecosystem. The sales presentation highlights modules such as EBK-4L4U-4C, EBK-8I8O-4COM-M, EBK-U8, EBK-QSFP28, EBK-UFS, and the newer EBK-210 functional safety extension.

Functional Safety Direction with T20-05

The newer RCB400 T20-05 and MARS400 T20-05 direction adds a functional safety focus based on NVIDIA Jetson IGX Thor T5000. The presentation links this variant with the EBK-210 Functional Safety MCU extension and references ISO/IEC TR 5469:2024 for AI systems and functional safety.

Software-Ready for AI and Real-Time Control

The product family is presented not only as hardware, but as part of a larger NEXCOBOT stack that includes NexRTOS and NexECM. That combination is designed to support AI inference and deterministic EtherCAT-based motion control on the same platform.

Best-Fit Applications

  • Humanoid robots: Thor-based MARS400 and RCB400 models for maximum AI performance and multi-function integration.
  • Industrial robot arms: AGX Orin and Thor platforms for motion control, vision, and high-speed communication.
  • AMRs and mobile robotics: Orin NX and Orin Nano variants for compact deployment, camera support, and rugged connectivity.
  • Factory automation: MARS400 system models for rapid deployment with industrial enclosure and broad I/O.
  • Safety-oriented AI machines: T20-05 with EBK-210 for functionally safer future robotics architectures.

Why the Series Matters

The strength of the RCB400 and MARS400 family is that it gives OEMs and system integrators one scalable hardware path instead of separate platforms for entry, mid-range, and flagship robotics systems. That reduces redesign effort, shortens product migration cycles, and helps standardize development across multiple machine classes.

From a commercial perspective, the product story is also clear: RCB400 is for customers building their own hardware around an embedded AI core, while MARS400 is for teams that want a ready industrial system with faster deployment and less mechanical engineering overhead.


Contact us!