The aim of the project is to supply the maintenance tracks with material from the warehouse using automated systems. To this end, InSystems has developed an automated guided vehicle system that fits into the complex environment of the ICE plant. The company opted for the proANT AGV 576 primarily because it was no longer able to maintain and find enough skilled workers for the internal material flow supply. The new technology enables fully autonomous delivery to the track aisles, regardless of future staff fluctuations. Employees who were previously responsible for transportation can now be deployed for other value-adding tasks. The lithium iron phosphate LiFeYPo4 batteries integrated in the transport robot enable smooth operation for up to six hours when fully charged. The lithium iron phosphate is non-toxic and non-flammable. The vehicle is charged at a 50A charging station. This has a drive-charge factor of five - for every hour that the transport robot charges, it can drive for five hours. The AGV manages the transport over the approximately 450 meters about seven times per hour. It travels fully automatically with the elevator from the basement to the maintenance level to bring the material from the material store to the train.

From the "ant" to the AGV
The ICE plant at Munich Central Station was gradually put into operation between 1989 and 1995. On a total area of around 36,000 square meters, there are six hall tracks for carrying out operational maintenance work on the ICE multiple units. The work can be carried out ergonomically on three working levels. Previously, hand pallet trucks were mainly used for transportation, or pallet trucks powered by fuel or lead-acid batteries were used. The distance between the material store and the material stations on the track is between 30 and 450 meters.

Orientation thanks to a "virtual red thread"
The autonomous driverless transport system is designed to optimize intralogistics in the plant. In an initial configuration phase, the modern pallet truck is moved manually using a built-in joystick. All the distance profiles recorded by the laser scanners are recorded and combined with the data from the motor encoders to create a map of the environment that shows the surfaces of the surrounding objects.
On the basis of the map created, the route planning is set using an easy-to-use and flexible visualization. Within the pre-scanned map, the robot is given a virtual line that it follows and does not leave during subsequent operation. Various action commands can be set at defined nodes, so-called goals, such as setting down load carriers, opening doors or checking the status of an intersection. Once the robot is in operation, it starts an initial localization process to determine its position. To do this, it compares the distance profiles detected by its laser scanners with the corresponding parts of the map. This localization algorithm takes place at regular intervals during autonomous driving operation so that the robot can constantly verify its position and navigate reliably.

Sensors help with space selection
In the material warehouse in the basement of the ICE plant, a picking area with five load carrier storage locations is created and managed. The storage locations are marked by yellow lines on the floor and are used to provide pallets for transportation and to store pallets after return transportation. The transport orders are generated using a graphical user interface. In the track aisles on the maintenance level, three load carrier spaces are also set up at each of five material stations. A sensor is installed above each parking space so that the robot automatically detects which space is occupied or free at the material station in the track aisle when pallets are delivered. When the vehicle delivers a pallet, the final location in the material yard is only determined shortly before arrival. The robot navigates in the taught-in map to a so-called "arrival selector", which is set up at each material station. Once there, the status of the sensors at the material station is evaluated in the fleet management server. The robot places the pallet in the first free space. If all the load carrier slots are occupied, the vehicle stops and reports an error until a slot becomes free.

Coupled: elevator and AGV
On the way from the basement to the maintenance level, the vehicle must use the existing elevator system. For this purpose, a coupling PLC was installed in the elevator control room as an electrical interface between the AGV and the elevator. The coupling PLC receives commands from the fleet management server and controls the elevator by setting certain signals, which in turn bypass the elevator buttons. In addition, a light was installed on the elevator to indicate to employees that a vehicle is in the elevator. If the elevator is reserved by a robot, the lamps flash until the vehicle has left the elevator again. The buttons for manually calling the elevator are blocked in the meantime. In addition to the transport robot, manually operated conveyor vehicles such as forklift trucks also operate on the maintenance level. If the travel path is blocked by a forklift truck, a command button on the proANT and in the elevator can be used to instruct the vehicle to move to a parking position in order to clear the travel path.

Scalable solution for dynamic environments
The autonomously navigating transport robot handles the transportation of pallets completely autonomously, meaning that no employees need to be involved in transportation tasks. Another advantage: the AGV can be installed without changes to the infrastructure and navigates in a dynamic environment. The solution helps to minimize bottlenecks in the process supply, all material movements are logged and the material flow is therefore more transparent. In addition, the fleet of transport robots is scalable for future increases in production.