InWareDrones, inventory by using drones and RFID

Over the past two years, a group of technology partners (Imec, DroneMatrix, ID Logistics Belgium, In2power, Octinion, Flanders Make and Aucxis) has been working on an autonomous drone inventory solution for high-bay warehouses. The drone operates completely independently: it’s flying around, avoiding obstacles, reloading itself,… All without human intervention.

In addition to the overall project management in this collaborative research project InWareDrones, Aucxis was responsible for the RFID part of the solution.

Stock control in department stores and warehouses is often still performed manually, which makes it a very labour-intensive task. In order to make this time-consuming and error-prone process more efficient, automated systems are required, especially in high-bay warehouses. The InWareDrones research project team developed a drone that can fly autonomously through the racks to check the stock.

Given the high reading speed, RFID is an interesting technology to be implemented in this drone solution. Aucxis focused on building a compact and lightweight RFID solution for the identification and localisation of pallets.

Pieter Suanet, Project Lead InWareDrones and R&D Engineer at Aucxis: “In the end, the drone system has both RFID and vision (4 cameras) technology on board. The RFID solution consists of two antenna arrays  enabling to emit RFID radiation in predefined directions, the so-called ‘beam steering’.”

“At the end of the research project, the accuracy of the exact localisation by means of RFID turned out to be roughly between 80-95%, and the inventory as such at 99.9%,” he continues. “The exact result mainly depends on the location of the RFID labels: the more centrally they are placed on the pallet, the better”. 

In addition to identification, also a large number of sensors and technologies are on board to enable the drone to navigate accurately and safely through the high-bay warehouse and land and take off autonomously, namely SLAM, UWB, ArUco markers, LiDar, VLP and sensor fusion.

Imec: “The main goal of the project was to reduce inventory management costs. A solid autonomous stock management system saves the customer hundreds of working hours, reduces the number of incidents and provides qualitative stock information at a much higher frequency. Thanks to the collaboration between researchers and industrial partners, we can conclude this has been a successful project”.