On the path to autonomous outdoor operation of forklift trucks

KION subsidiary Linde Material Handling and the Aschaffenburg University of Applied Sciences presented the results of the research project “KAnIS – Cooperative Autonomous Intralogistics Systems” with live demonstrations on the test site at the Linde plant in Aschaffenburg. In several subprojects, solutions were developed for the demanding applications of autonomous counterbalanced forklifts, which transport loads both indoors and outdoors. One focus is on the cooperative behavior of these vehicles that exchange information in real time via a 5G network and an edge server and can warn each other of obstacles.

“Thanks to the joint research work with Aschaffenburg UAS, we have been able to develop viable solutions for these complex requirements. Once the project is completed, these findings will form an essential basis for further development projects,” says Stefan Prokosch, initiator of the KAnIS project at Linde MH.

The overall goal of the project was to investigate how the cooperative behavior of networked, autonomous vehicles can improve operational reliability and handling performance. To solve this broad task, several subprojects were formed to address vehicle location, regulation and control as well as forklift cooperation, load carrier recognition, the impact of weather influences, predictive maintenance, route optimization and automatic load management.

“For the university, the KAnIS project was a very complex, interdisciplinary research project. Ten professors and numerous research assistants and students were involved,” says Prof. Dr. Hans-Georg Stark, Project Manager KAnIS, Faculty of Engineering at Aschaffenburg UAS.

Four Linde E20, E25 and E30 electric counterbalanced trucks with a load capacity of 2.0 to 3.0 tons were automated and equipped with electrohydraulic steering (Linde Steer Control), the Linde Safety Pilot assistance system with electronic load diagram and an integrated fork positioner. Starting next year, the vehicles are to be further developed and tested so that they can perform four specific material handling tasks in the future. These include the transport of wire mesh crates and of pallets containing batteries, and the relocation of vehicle frames and overhead guards, which have to be transported on special load carriers from pre-assembly to the main assembly lines. The first two applications are purely outdoor operations, while the other two require the trucks to travel both inside and between the halls. Gradients of 8 percent have to be overcome, and there are also other AGVs and manually operated vehicles in the halls.

A particular focus of the research project was on the automated forklifts’ perception of their surroundings in order to ensure their reliable interaction with other road users. For this purpose, the vehicles are equipped with 3D scanners and HD cameras in addition to the sensors of the personal protection system. The camera data forms the basis for detecting and classifying objects with the help of AI algorithms and then locating them in order to adjust the vehicle’s speed and slow it down to a standstill. But that’s not all. Another key issue focused on critical situations that arise when people are in concealed areas that cannot be detected by the forklift’s sensors and approach the vehicle’s path of travel. This is where cooperation between the forklift trucks comes into play, because if another forklift is in the vicinity, it can provide the relevant information. However, this requires real-time transmission of the perception data. To achieve these low latencies, Linde has set up a private 5G network at the Aschaffenburg plant. The perception data is transmitted from the forklifts to an edge server, which uses the locally detected objects to create a global list of all detected objects and sends it back to the forklifts.

The test was conducted using a crash-test dummy that suddenly emerges from behind a wall and runs into the forklift’s path. Without cooperative behavior, the automated forklift truck cannot stop in time and runs into the dummy. However, if it receives real-time information from a nearby forklift, it can anticipate the dangerous situation in advance and brake in time. Since it is not always possible to assume that a second forklift is nearby, eight stationary 3D laser scanners were installed at intersections and gateways along the routes that the KAnIS forklift trucks will travel in the future. The local object lists of the stationary laser scanners are also merged on the edge server and the information is made available to all vehicles.