Motion analysis platform selected as finalist in AI and Robotics Sprint
Traversal Labs has been selected as a finalist in the inaugural Propel-AIR program, Australia’s first dedicated AI and Robotics Sprint, with its MotionLogic platform.
While the world’s largest tech companies pour billions into building humanoid robots, Newcastle-based Traversal Labs has taken a contrarian approach: it is teaching machines to understand human work patterns first. The MotionLogic platform transforms ordinary factory video into detailed maps of human motion and interaction in order to create precise training data that future robots will 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.
MotionLogic is already deployed with 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 is designed to deliver value through process optimisation. 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.
The platform's multi-layered approach captures three distinct layers. The person layer tracks ergonomics and body movement. The tool layer monitors interaction timing. The robot layer maps motion patterns and environmental interaction. This creates a comprehensive digital twin of factory operations, helping future robots get better at understanding, learning from and collaborating with human workers.
McClung, a robotics expert with over a decade building autonomous systems, founded Traversal Labs in 2021 after receiving a perfect score on his honours thesis in visual AI. The company represents three years of focused development in spatial intelligence.
“MotionLogic turns factories into data-rich learning environments,” McClung said. “This sensing capability is critical for the next generation of robotic co-workers.”
The software is applicable to sectors with complex assembly processes, such as shipbuilding, train assembly and mining maintenance. In these industries, motion inefficiencies translate directly to major operational costs.
Through the Propel-AIR program, delivered by ARM Hub in partnership with the Australian Government’s AI Adopt Centre, MotionLogic will compete against five other Australian finalists for the ultimate prize. The winner receives a four-week residency at MassRobotics in Boston, providing exposure to global leaders including Boston Dynamics and iRobot, plus access to MIT and Harvard.
The program concludes with a virtual pitch event on 31 July, with the winner announced on 4 August.
Konica Minolta and MiR launching two autonomous robots at CeMat
Konica Minolta Australia has announced that it will be launching two autonomous mobile robots...
Australian researchers create 'cyborg' beetles
Researchers at UQ are experimenting with 'cyborg' beetles that they hope could...
Siemens to add AI and robotics capabilities to AGVs
Siemens has announced GenAI and robotics capabilities for better management and safety of...