View organization page for Flowcate

2,100 followers

Tech Tuesday: What humanoid robots can learn from location tracking   Humanoid robots and AMRs are moving into warehouses, factory floors, hospitals, and many other areas. How confidently they move depends on one thing: knowing exactly where they are, at all times.   No single positioning technology delivers that everywhere. UWB, BLE, Wi-Fi, GPS, RFID, and vision each work well in some environments and poorly in others. A robot might get a precise UWB fix indoors while at the same time receiving a low-accuracy GPS update. So the real question isn't "where am I?" but "which of these signals do I trust right now?"   That's exactly what the DeepHub rule engine solves. Instead of relying on one technology, it evaluates every incoming location update in real time and picks the most significant one based on configurable rules. Each rule combines a condition and a priority. A simple example: use the precise UWB position as long as it's less than 10 seconds old, otherwise fall back to GPS if its accuracy is better than 10 meters. Rules can be built on properties like technology type, accuracy, the age of the fix, speed, floor, or zone, so the best available source is defined by the situation rather than by a fixed hierarchy.   The result is a robot that switches positioning sources automatically and seamlessly as it moves through changing environments. We contributed this capability to the omlox standard, so multi-technology tracking becomes an open, interoperable foundation instead of a proprietary feature.   And where robots operate close to people, the engine can run several technologies in parallel for redundant tracking. That means reliability exactly where it matters most.   Robots make the headlines. The positioning layer underneath is what has to work. #Flowcate #omlox #Robotics #DeepHub #RTLS

To view or add a comment, sign in

Explore content categories