As someone deeply involved in optimizing industrial automation solutions, I’ve been exploring the potential of OpenThread for location inference in large-scale warehouses. The warehouse I’m working with spans 110 meters by 25 meters, divided into 16 zones, each housing between 30 to 55 devices. These devices wake up every 5-10 minutes to transmit sensor data, which is then used to dynamically adjust ambient conditions.
The challenge lies in accurately determining the location of each device within this vast space. Traditional methods like BLE or LoRaWAN gateways become impractical due to the warehouse’s size and the high cost of infrastructure. This is where OpenThread comes into play. It offers a promising alternative by leveraging the network topology to infer device locations.
My proposed approach involves each end-device connecting to the nearest router, which would then map the device’s location based on the router’s MAC address. However, initial tests revealed that devices don’t always connect to the nearest router, likely due to OpenThread’s self-balancing mechanism prioritizing hop counts over physical proximity. To address this, I’m considering increasing device density per area or reducing transmission power to encourage connections to closer routers.
While I’m still in the testing phase, I’d love to hear from others who have experience with OpenThread in large-scale environments. Any insights or alternative approaches would be invaluable as I continue to refine this solution. The potential to streamline warehouse operations with precise location data is immense, and I’m excited to explore how OpenThread can make this a reality.