Forklift trucks and pallet trucks are part of everyday life in the intralogistics of many companies when loading and unloading trucks with pallets. The optimization of warehouses and many storage and retrieval processes are hardly imaginable without these powerhouses. As a rule, such vehicles are still operated manually. With a system for automated pallet recognition based on time-of-flight cameras, ifm electronic aims to make such processes more efficient, reduce the error rate compared to manual operation and improve the overall equipment effectiveness of companies. The aim is to increase productivity in logistics processes and improve the effectiveness of automated guided vehicles (AGVs).

Some of these autonomous vehicles are already equipped with systems that allow them to recognize the pockets in which the forks of the transport vehicles need to be positioned. Laser lines are usually used in these cases, but they require a relatively long time for reliable position detection. Developers at ifm robotics have therefore taken a closer look at this task and created a system called 'Pallet Detection System', or 'PDS' for short, based on the company's own time-of-flight cameras.

Time-of-flight as a basis

PDS' is based on the time-of-flight cameras from ifm's O3D series. They work according to the time-of-flight method and use the phase shift of several light pulses that are emitted and reflected by objects. By measuring the time delay between the emitted light pulse and the reflected light pulse detected by a sensor, the distance from each reflecting object point to the sensor can be determined relatively precisely. With this technology, it is therefore possible to measure the distances of each individual object point to the sensor and thus take three-dimensional images of objects.

The pallet detection system identifies the exact position of all standard pallets with two pockets. The system is based on time-of-flight cameras from ifm's O3D series, which are typically installed slightly above and between the forks of the vehicles.

© ifm

The developers used this property for the pallet recognition system: they installed a ToF camera slightly above and between the forks of an autonomous forklift truck, which moves to a distance of between one and two meters from the next pallet to be gripped during operation. Using the triggers, the ToF camera records a complete 3D point cloud of the pallet, which is then filtered using the 'PDS' algorithm to eliminate noisy or unnecessary pixels. As a result, the system outputs the pallet's degrees-of-freedom pose, which displays the exact pallet coordinates in the x, y and z directions as well as any rotations around the vertical axis or horizontal tilts. The 'PDS' communicates this data by wire via an Ethernet interface or a CAN field bus to the vehicle's control system, which then ensures that the load is picked up correctly.