Video: Extend Robotics Tests VR-Controlled Humanoid Robot for Hazardous EV Truck Tasks at Leyland

Extend Robotics Trials VR-Controlled Robot for Hazardous EV Truck Tasks at Leyland

UK-based Extend Robotics has partnered with commercial vehicle manufacturer Leyland Trucks to test the feasibility of using teleoperated robots for potentially hazardous tasks involved in electric vehicle (EV) assembly. Recent trials at Leyland's facility focused on using a robot controlled via a virtual reality (VR) interface to handle high-voltage components, aiming to improve worker safety and operational efficiency.

The Challenge: High-Voltage Safety

The specific task demonstrated involved connecting the Master Service Disconnect (MSD) unit on EV trucks. According to Leyland Trucks personnel involved in the trial, this procedure currently requires human operators to don extensive personal protective equipment (PPE), including visors, head protection, neck guards, and specially tested gloves. The process of getting fully suited up reportedly takes around 20 minutes, highlighting the time overhead and inherent risks associated with working near high-voltage systems.

Leyland Trucks, a subsidiary of PACCAR Inc, sees robotic assistance as a potential solution to mitigate these dangers and streamline production. The company aims to increase its output from approximately 80 trucks per day towards over 100, and leveraging automation is part of that strategy.

The Solution: VR Teleoperation and AI Learning

Extend Robotics' platform allows operators to remotely control a robot using an intuitive VR/XR (extended reality) headset and controllers. Video footage from the trial shows an operator guiding the robot's arm and hand to successfully pick up and install the MSD unit. Workers participating in the test, including one with no prior VR experience, described the controls as simple and natural, remarking, "It's like your hand is that hand."

Beyond direct teleoperation, Extend Robotics is developing AI capabilities designed to allow robots to learn tasks by observing expert human operators. During the trial, a basic AI function was demonstrated where the robot could autonomously identify and pick up a green object from a cluttered selection, ignoring red ones. The goal is for the robot to eventually learn and perform complex tasks autonomously after being trained through demonstration.

Early Results and Future Potential

The initial two-day feasibility study yielded positive feedback. Leyland's Innovation Manager, Adrian Walsh, commented on the "good early quick success" and noted, "I think we've got a great user case to continue here."

The collaboration represents a step towards integrating advanced robotics into complex manufacturing workflows where full automation is challenging, but human safety is paramount. By combining human oversight via VR with increasingly capable robots that can learn on the job, the approach could offer a scalable path to automating dangerous or intricate tasks.

While still in the exploratory phase, the trials at Leyland Trucks highlight the growing potential for teleoperated and AI-enhanced robots to work alongside humans in industrial settings, tackling jobs that are too risky or ergonomically challenging for people alone.