The vehicles generally move through the factory on fixed routes and always perform the same tasks fully automatically. However, as with any technical system, AGVs also occasionally experience malfunctions. For example, if an obstacle blocks the path and the vehicle cannot drive around this obstacle without leaving its predefined route. Most automated guided vehicles then stop until the obstacle is removed - or until they are manually steered around the obstacle.

With large transport vehicles such as forklift trucks, it is sometimes possible for a person to simply get behind the wheel, provided an employee with the appropriate driver's license is nearby. This is not possible with smaller transport units. They could be moved around the obstacle manually, but this can lead to a safety risk for the employee and damage to the vehicle. For this reason, manual intervention is usually carried out via the control system.

The control system determines which vehicle completes which jobs on which route and in which order. To intervene here, specialist knowledge and special access rights are required, which for security reasons are normally only held by one or two employees in the company. In small companies, it may even be the case that no one has access to the control system and the manufacturer of the AGV has to be called every time there is a fault. This is time-consuming and cost-intensive.

Scientists at the Institut für Integrierte Produktion Hannover (IPH) gGmbH have developed an intuitive and mobile system for human-machine interaction as part of the MobiMMI research project in order to be able to control vehicles at any time - without special training, without IT knowledge and without interfering with the control system. The system is not yet available to buy, as the scientists still need an industrial partner to develop it to market maturity together with them. However, the demonstrator created in the research project can be tried out at the IPH.

From the outside, the MobiMMI demonstrator is nothing more than a standard pair of AR glasses. What makes it special is the software, which is the result of two years of research and development. This software can be used to control automated guided vehicles in a mobile and intuitive way. Mobile means that the employee can stand anywhere in the factory within sight of the vehicle instead of sitting far away at a computer. "Intuitive means that the employee communicates with the vehicles in the same way that people do: via eye contact, speech and gestures," explains MobiMMI project manager Andreas Seel. The logistics employee simply puts on the AR glasses and can then have a three-dimensional joystick displayed. This allows them to steer the vehicle around obstacles or to stations that are not programmed, such as chaotically parked deliveries in the incoming goods area, with just a few hand movements. Manual control can also be useful for non-routine transport orders. "Driverless vehicles may only transport goods that do not exceed certain dimensions. In the case of oversized or bulky goods, a logistics employee can control the vehicle manually and make sure that it doesn't hit anything," says Andreas Seel.

As an alternative to gesture control, the employee can also activate voice control, select individual automated guided vehicles (AGVs) and specify a destination, for example: "AGV 3, go to station 5." Voice control is suitable for unscheduled transport orders. Normally, AGVs work through one programmed job after another. If something urgent needs to be done, a qualified employee has to intervene via the control system and bring the job forward. Thanks to the software in the smart glasses, the MobiMMI system can do this on demand. "Just like I would shout to the forklift driver in the factory: Can you just pick up this pallet? I can now also give orders to automated guided vehicles," explains Florian Kreutzjans, who was responsible for development in the MobiMMI research project and is presenting the demonstrator at IPH.