Physical AI: R&D Hub
Physical AI is new ground, and we treat it that way: as a serious R&D investment rather than a finished product. Our R&D hub is taking the spatial and edge-vision engineering we've proven elsewhere into one of the hardest operating environments on earth, the container terminal, working toward crane automation, yard intelligence and lift-prevention safety.
We're honest about where this sits. The direction is deliberate and the underlying engineering is real, but this is active research and development, built with port operators rather than sold as a packaged solution.
What we build here.
Crane automation research
Vision-assisted spreader positioning and container handling on the quay and in the stack.
Yard mapping & profiling
Centimetre-grade positional awareness of stacks, lanes and equipment in real time.
Lift-prevention safety (CLPS)
Detecting and stopping accidental chassis lifts, a safety-critical edge-vision problem.
Auto gantry steering (AGSS)
Keeping RTG/RMG cranes tracked precisely, automatically, run after run.
Edge computer vision
On-device inference where latency and connectivity rule the cloud out.
World-model exploration
Tracking and prototyping with the physical-AI and world-model research reshaping the field.
Proof, not promises.
Physical AI for the terminal
An open frontier we're researching with port operators. The spatial command and edge-vision engineering behind it is proven in production elsewhere; our R&D hub is now adapting it toward crane automation, yard profiling and edge-vision safety on the quay.