Advantech and Aeolus Robotics Enable Edge AI Humanoid Robots with AFE-R750

An edge AI humanoid service robot powered by Advantech AFE-R750, NVIDIA® Jetson AGX Orin™, Advantech Robotic Suite, and Aeolus Robotics’ service robot software to accelerate real-world humanoid deployment.

Taipei, Taiwan, 26, May, 2026 - Advantech (TWSE: 2395), a global leader in AIoT and Edge Computing, today announced a technology collaboration with Aeolus Robotics to advance next-generation humanoid service robots for real-world indoor applications. Built on Advantech AFE-R750 and powered by NVIDIA® Jetson AGX Orin™, the joint solution integrates NVIDIA® Isaac ROS, Advantech Robotic Suite, and Aeolus Robotics’ service robot software to enable natural language interaction, robotic arm control, environmental awareness, and autonomous operation at the edge. By combining industrial-grade computing, validated robotics frameworks, and application-ready robot intelligence, the collaboration provides a scalable foundation for bringing humanoid service robots from prototype to practical use.

As humanoid robots evolve from demonstrations into commercial applications, developers face growing challenges in system integration, AI performance, sensor fusion, software compatibility, and long-term reliability. Advantech and Aeolus Robotics are working together to address these challenges with an integrated architecture that connects edge AI computing, robot control, software development tools, and service-oriented applications.

This approach helps reduce engineering complexity while supporting faster development of robots that can interact naturally with people and operate reliably in dynamic indoor environments.

At the core of the solution is Advantech AFE-R750, a robotics-ready control system powered by NVIDIA Jetson AGX Orin. Designed for intelligent mobile robots and humanoid service robots, AFE-R750 delivers high-performance AI computing, rich I/O connectivity, and stable system operation in a compact industrial-grade design. With the ability to support multimodal AI workloads, sensor input, and robot control functions, AFE-R750 serves as the robot’s central computing foundation for advanced service applications.

The solution also incorporates NVIDIA Isaac ROS and Advantech Robotic Suite to streamline robotics development. Through ROS2-based software building blocks, AI frameworks, sensor integration, and validated workflows, developers can simplify system bring-up and shorten the path from proof of concept to field-ready solutions. This software foundation allows robotics teams to focus more on user experience, application logic, and service innovation rather than low-level integration tasks.

Aeolus Robotics brings proven expertise in service robot development, combining in-house robot design, AI software, human-robot interaction, and deployment-ready robot management. Together with Advantech’s AFE-R750, Aeolus Robotics’ software enables humanoid robots to conduct multilingual conversations, respond to user interactions, perform robotic arm movements, and support service workflows in real-world indoor scenarios.

“Humanoid service robots require more than advanced mechanical design. They need a reliable AI ‘core brain’ that can process vision, language, motion, and decision-making in real time,” said Shaubin Xie, Software Director at Aeolus Robotics. “By working with Advantech, we are able to combine Aeolus Robotics’ full-stack service robot software with an industrialgrade edge AI platform, enabling robots to interact more naturally with people and operate more reliably in dynamic indoor environments.”

“The evolution of humanoid robots depends on continuous advances in AI computing, perception, motion control, and system integration,” said Aaron Su, Vice President of Embedded Sector at Advantech. “Through our collaboration with Aeolus Robotics, Advantech delivers the AFE-R750 robotics-ready platform and Robotic Suite software services, together with Aeolus Robotics’ expertise in service robotics and application software. We are accelerating the deployment of smarter, more responsive, and more reliable humanoid robots.”

The joint solution demonstrates how edge AI, robotics software, and system-level integration can work together to support the next generation of service robots. By combining Advantech’s industrial-grade robotics platform, and Aeolus Robotics’ full-stack robot application expertise, the collaboration provides a practical path toward more intelligent, responsive, and commercially deployable humanoid service robots.

Advantech and Aeolus Robotics will showcase the joint humanoid service robot solution at COMPUTEX, where visitors can experience live chatbot interaction, robotic arm movement, multilingual conversation, product introduction, and real-world deployment video demonstrations.

Learn more about AFE-R750: here 

Learn more about Advantech Robotic Solutions: here

About Advantech

Advantech is a global leader in IoT intelligent systems and embedded platforms, with the corporate vision of "Enabling an Intelligent Planet." To embrace the trends of edge computing and artificial intelligence, Advantech fully deploys its sectordriven strategy and focus on Edge Computing and Edge AI. In the meantime, Advantech is enhancing its global presence and core competitiveness by integrating its Edge Computing hardware platform, the WISE-IoT software platform, and sector-specific Edge AI solutions with domain expertise.

About Aeolus Robotics

Aeolus Robotics develops advanced AI-powered service robots that combine autonomous navigation, multi-sensor perception, and dual-arm manipulation to support real-world commercial and care environments. Its flagship robot, Aeobot, is designed to assist with tasks such as eldercare monitoring, security patrol, facility operations, disinfection, guidance, and multi-floor mobility. By enabling robots to interact with existing doors, elevators, and physical infrastructure, Aeolus Robotics helps organizations address labor shortages, improve operational efficiency, and bring practical embodied AI into everyday service workflows.