While Tech Giants Build Humanoid Robots, This Australian Startup Solves the Harder Problem First
MotionLogic Selected as Finalist in Australia’s First AI and Robotics Sprint NEWCASTLE: Traversal Labs has been selected as a finalist in the inaugural Propel-AIR program, Australia’s first dedicated AI and Robotics Sprint, with their MotionLogic platform. While the world’s largest tech companies pour billions into building humanoid robots, Newcastle-based Traversal Labs has taken a contrarian approach. They’re teaching machines to understand human work patterns first. Their MotionLogic platform transforms ordinary factory video into detailed maps of human motion and interaction. This creates precise training data future robots need to work safely alongside people. “Everyone’s building the robot body,” said Alex McClung, founder of Traversal Labs. “We’re building the robot’s understanding of how humans actually work.” The AI-powered platform analyses camera footage to extract key performance data. It measures motion waste, task breakdowns, safety risks, and ergonomics. Workers’ movements are tracked in 3D space, measuring everything from walking distance to joint angles and tool handling efficiency. Australian Innovation Gains Industrial Traction MotionLogic is already deployed with enterprise clients in the heavy-manufacturing sector, with negotiations underway in the Aerospace, Rail, and Maritime sectors focusing on customers with complex assembly processes. The technology delivers immediate value through process optimization. A typical deployment reveals insights like total walking distance per shift, frequency of awkward postures, and opportunities for improved factory layout. The software provides production staff with detailed visibility into when tools are active versus idle and when workers are engaged in task-time versus non-task-time. But the long-term value lies in robotics training data. “Just as self-driving car companies use real-world driver data to train autonomous vehicles, MotionLogic feeds real-world human work patterns into industrial robot training,” McClung explained. From Ergonomics to Automation The platform’s multi-layered approach captures three distinct layers. The person layer tracks ergonomics and body movement. The tool layer monitors interaction […]