More than 1,000 flight hours in prototype mode across three pilot warehouses were required to arrive at this product, which allows companies to perform completely automatic warehouse inventories without interrupting their usual operations or requiring any human intervention. The main advantages are the productivity gains generated by performing the inventories outside warehouse operating hours, greater safety at work for the site’s employees, who no longer have to carry out this tedious and sometimes risky task, and a greater reliability of the inventory.
In the prototyping phase, multiple tests were carried out in real operating conditions in order to meet the specific constraints of warehouses, such as low light or the impact that the plastic film covering pallets has on image capturing, which required special adaptations to the cameras embedded in the drones.
This phase ended in 2017 with the development of a complete ‘plug and play’ system that can be easily moved from one warehouse to another without requiring that any prior changes be made to the warehouse. It also adapts to all types of warehouse management systems (WMS).
Based on these positive results, GEODIS and Delta Drone are now working on producing a system that can be manufactured, focusing their attention on the solution’s design and the final choice of the best components. The aim is to move to the industrial production stage at the end of the year. Initially, GEODIS will be the exclusive user of this product in its own warehouses.
The system designed by the GEODIS and Delta Drone engineering teams is unique. It combines a ground-based robot equipped with a battery that provides the energy needed to navigate a warehouse and allows freedom from autonomy constraints, and a quadcopter drone equipped with four high-definition cameras. This set, equipped with indoor geolocation technology, operates autonomously during the hours the site is closed. From an IT point of view, the set-up enables the counting and reporting of data in real time, the processing of data, and its restitution into the warehouse’s information system. The tests conducted during this initial development phase show that the system enables inventory to be managed reliably with rates close to 100%.