(MENAFN- Robotics & Automation News) Interview with NexCOBOT's general manager: 'Single control platform'

October 14, 2025 by Sam Francis

As humanoid robots move from research labs into production environments, one of the biggest challenges for developers is system integration.

Building a robot that can see, move, and interact safely with humans requires multiple layers of computing – AI inference, real-time motion control, and certified functional safety – traditionally managed on separate platforms.

Bridging those systems efficiently is becoming the defining test for the next wave of humanoid design.

Few companies sit at that intersection as clearly as NexCOBOT , a subsidiary of NEXCOM International , which has long specialized in industrial PC architecture and automation controllers.

Headquartered in Taipei, NexCOBOT focuses on open, modular control platforms for robotics and AI applications.

Under the leadership of general manager Jenny Shern , the company has built a strong reputation for applying industrial-grade computing discipline to emerging fields such as collaborative and humanoid robotics.

Earlier this year, NexCOBOT attracted attention when Nvidia named its parent company, NEXCOM, as the only partner developing humanoid robots on the Jetson Thor platform.

The collaboration merges Nvidia high-performance edge AI computing with NexCOBOT's real-time EtherCAT master control software, creating an integrated control stack that can handle motion, safety, and AI inference on a single system.

The company's latest product, the MARS400 T10, consolidates these capabilities into a compact, TÜV-certified module aimed at humanoid robot builders.

In this Q&A, Jenny Shern discusses how NexCOBOT's partnership with Nvidia is reshaping control architecture for humanoid robotics, the technical and safety hurdles her team has addressed, and where she sees the next wave of humanoid deployments emerging – from manufacturing and logistics to public service and healthcare.

Interview with Jenny Shern, general manager of NexCOBOT
Jenny Shern

Robotics & Automation News: Partnership with Nvidia – NEXCOM is described as the only partner developing humanoid robots on Jetson Thor. What makes this collaboration unique, and how does it position NEXCOM in the broader humanoid robotics market?

Jenny Shern : NEXCOM stands out in the humanoid robotics space because of its comprehensive modular solution for builders, which includes a robotic AI controller, real-time motion controller (with real-time NexRTOS and NexECM EtherCAT Master), safety controller, robotic actuators and safety sensors.

Usually, robotic AI controls, real-time motion controls and functional safety controls are handled on three separate platforms. Through our partnership with Nvidia, we are able to apply our real-time EtherCAT master control software onto Nvidia's Jetson Thor platform.

This allows us to consolidate these distinct functions into a single control platform, which is ideal for the unique challenges of humanoid robot control.

In doing this, NEXCOM allows humanoid robot developers to shift their focus from system-level integration to building innovative applications, positioning us at the forefront of the evolving humanoid robotics market.

R&AN: Technical integration – The MARS400 T10 combines AI computing, motion control, and functional safety in a single system. What challenges did you face in merging these traditionally separate functions, and what advantages does this bring to end users?

JS : Maintaining real-time performance for EtherCAT applications is an important technical challenge when merging AI computing, motion control and functional safety into a single system like MARS400 T10. Any jitter in the system could compromise the robot's stability, accuracy and safety when interacting with humans.

Based on our experience, factors like IPC hardware, BIOS configuration, RTOS performance and the EtherCAT master implementation all contribute to these issues. The integration of diverse AI models only adds to the complexity, making it even more difficult to ensure a stable and safe robotic system.

As AI-driven humanoid robot use cases evolve, new safety standards and rules must be adopted based on specific applications and associated risk assessments.

However, the integrated design of the MARS400 T10 offers key advantages to end users, including reduced cabling and improved spatial efficiency – an essential benefit for compact humanoid robot bodies.

R&AN: Safety certification – NEXCOM is the only TÜV-certified provider of modular robot safety technology in Taiwan. How does this certification influence customer trust and adoption in industries that are cautious about deploying humanoids?

JS : NexCOBOT has been developing robot controllers for 10+ years and has been undergoing functional safety product certification with TÜV Rheinland since 2020.

Throughout the entire development process, from design planning and specification formulation to development and final product testing, TÜV Rheinland's professional assessment experts review the documentation to ensure compliance with worldwide functional safety requirements and standards.

At NexCOBOT, we earn our customers' trust through the careful efforts of our in-house hardware and software R&D team, which is dedicated to instilling functional safety standards in all their work.

Developing our safety controllers with these high standards has accelerated humanoid robot development and adoption across the industry.

With TÜV Rheinland-certified FuSa training and deep expertise in FSoE (FailSafe over EtherCAT) master-slave communication technology, our team develops all hardware and software in-house to ensure quality, reliability, and compliance.

R&AN: Industry applications – Beyond the technology showcase, what practical use cases or industries do you see adopting humanoid robots powered by the MARS400 T10 in the near term?

JS : We will mainly see humanoid robots deployed in manufacturing and logistics use cases, but there's also the possibility of expanding into industries like public safety, healthcare, and service industries.

MARS400 T10 makes it possible for wheeled or legged humanoids to be safely implemented into these fields because it reduces cabling and complexity in robotics controls.

Humanoids currently use a high-level AI control and low-level motion control hybrid architecture that includes multiple platforms within it.

MARS400 T10 simplifies this by consolidating the workload required for both real-time control and AI applications like training and inferencing.

R&AN: Future roadmap – With Nvidia and other chipmakers accelerating into embodied AI, where do you see NEXCOM's role evolving over the next five years, and how important will humanoid robotics be in your strategy?

JS : The more we see AI integrated into physical agents, the more imperative safety and reliability become. To keep up with the fast-paced growth of the humanoid robot market and demand for human-robot coexistence, NEXCOM aims to develop an AI safety controller for humanoid robot developers that will follow the latest ISO/IEC TS 5469 standard for AI functional safety.

The system will use the latest Nvidia AI chip, combining motion control and AI safety computing into a single hardware platform capable of delivering up to 2070 TFLOPS of processing power.

The controller will also have a real-time operating system and an EtherCAT master, enabling fast and precise joint control while also running AI model inference such as VLA (Vision Language Action).

The safety EtherCAT master supports FSoE, allowing real-time monitoring of joint position, speed and force to ensure safe motion.

For AI safety functions, we will work with our partners to develop AI-based human detection and behavior monitoring technology, helping companies quickly apply it in real-world settings and improving the safety and reliability of human-robot interactions.

MENAFN14102025005532012229ID1110195614