MilliTrack AMR+ Help Robots Navigate Where Cameras And LiDAR Fail

The useful part

Autonomous mobile robots (AMRs) increasingly move materials around factories and warehouses, but most still rely heavily on cameras and LiDAR to understand where they are. Those systems can struggle in environments filled with dust, fog, excessive light, or long rows of nearly identical shelves. South Korean startup MilliTrack wants to address those blind spots with AMR+, an add-on localization system that uses millimeter-wave radio signals instead of relying solely on what the robot can see.

How it works

• Rather than replacing an AMR's existing navigation system, AMR+ supplements it.
• A tracker is installed on the robot, while small tags can be placed in areas where the robot's normal SLAM localization becomes unreliable.
• SLAM is a technology that lets a robot build a map of its surroundings while simultaneously figuring out where it is on that map.
• The system can then provide the AMR with another, better, source for its position.
• MilliTrack claims 3 cm localization accuracy at distances up to 30 meters, with latency below 100 milliseconds.

What to take from it

Importantly, AMR+ does not require the robot to abandon its existing LiDAR, cameras, or inertial sensors. During our conversation with MilliTrack CEO Song Min Kim, he explained that an AMR can monitor the confidence level of its normal SLAM localization and switch to coordinates provided by AMR+ when that confidence drops.

Details worth keeping

These are not rare occurrences at industrial locations. MilliTrack's system requires line of sight between the tracker and a tag.