Pick and Move

Whether in warehouse logistics or internal material transport: In the food industry, a wide variety of products, orders and stocks have to be processed. According to the International Federation of Robotics (IFR), a new generation of AI-supported robots is helping to overcome these challenges. To enable machines to assist with flexible pick-and-move scenarios, their developers use AI software that relies on an experience-based learning process rather than programming. For example, this enables the robots to pick and pack various items at high speed in a logistics centre.

Other innovations that have already been significantly improved through the use of AI include autonomous mobile robots (AMR) and automated guided vehicles (AGVs). "These systems navigate independently through warehouses, avoid collisions and continuously optimise their routes," says Rainer Schulz, Managing Director at Sysmat. "Thanks to the integration of sensor technologies and AI-based control systems, they work seamlessly with existing infrastructures and enable dynamic adaptation to changing production and logistics processes," explains the expert.

A current example of AI-supported order picking is the Flexley Tug T702 from ABB, the first of an entire series of mobile robots with Visual SLAM navigation (Simultaneous Localisation and Mapping). The autonomous mobile robot (AMR) uses AI-supported 3D image processing for positioning and mapping and is able to make intelligent navigation decisions based on its environment. Another example from ABB is picking modules for logistics and e-commerce applications that use intelligent image processing technology to handle unknown and randomly arranged items in unstructured environments. This ensures seamless processing with high throughput and a high mix of variants. The robots have a picking accuracy of over 99.5 percent – even in highly dynamic environments where item sizes, shapes and packaging types change daily. The motion planning software enables collision-free automatic route planning as soon as each item has been identified by the AI vision system. In addition, the AI system can also be trained to recognise and sort out non-separable items to ensure reliable, efficient operation.

Handling items and transporting them from A to B; the next revolution will be systems that can do both – at least if the developers at evoBOT have their way. What makes it special: The dynamic modular robot, equipped with gripper arms and running on two wheels, can be used not only as a transport robot, but also collaboratively as an assistant for people. To this end, its design principle is based on an inverted pendulum that does not require an external counterweight. Thanks to its pendulum motion, the mobile robot can lift objects directly from the floor and place them at different heights. In preparation for potential deployments in industrial settings, the Fraunhofer Institute for Material Flow and Logistics (IML) demonstrated its versatility to the public for the first time at Munich Airport. There, evoBot placed packages from a Euro pallet onto the conveyor belt of an X-ray machine and, after the inspection process, returned them to the pallet, while the omnidirectional, highly dynamic O³dyn robot, also developed by Fraunhofer IML, took over the transport of Euro pallets to the neighbouring warehouse. The processes were controlled using Fraunhofer's openTCS control system software – a low-threshold tool for coordinating automated guided vehicles (AGVs). "Artificial intelligence will support us in coordinating and controlling the vehicles in the future," explains Prof. Michael Henke. "It provides the necessary tools and algorithms that enable us to calculate the paths of the autonomous robots in advance and safely avoid collisions. In the end, we'll soon have completely autonomous systems," says the Managing Director of the Fraunhofer IML in summary.

The Fraunhofer Institute for Manufacturing Engineering and Automation (IPA) is also contributing its expertise in AI in intralogistics, focusing on the use of multifunctional and location-flexible robot systems. The researchers aim to create the conditions for this within the RoX project. The project, which was launched in September 2024, is intended to create a digital ecosystem for AI-based robotics applications, particularly in production and logistics. "For Germany as a production site and home to many of the world's leading suppliers of robot components and solutions, the project is giving impetus to new standards in a fast-moving international industry," says Richard Bormann, Research Team Leader and Project Coordinator at the Fraunhofer IPA.

The Fraunhofer IPA is contributing numerous technological developments to the project for a wide range of application contexts, which also play a central role in the food industry:

Flexible gripping solutions: This involves the robot-based handling of a wide variety of objects, for example in intralogistics and production. With the help of AI technologies, it will be possible to reduce set-up times and make robots capable of learning, for example with regard to necessary movement sequences or gripping techniques. These will then no longer need to be extensively programmed.

Robot-based intralogistics processes: The focus here is on loading and unloading trucks using autonomous mobile robots or autonomous forklifts, as well as on external transport within the factory, for example between factory buildings.

AI-supported design and commissioning: Based on the Computer-Aided Risk Assessment (CARA) software tool and the Robo-Dashcam, a camera for detecting people in the robot's environment, safe, performance-optimised applications can be implemented in a semi-automated manner. This simplifies and accelerates the currently complex risk assessment process, particularly for applications in which interaction between humans and robots is desired.

If the use of robotics in logistics and transport continues to increase, this will ease the burden on food producers who are unable to find personnel for these tasks. Above all, the introduction of a new generation of AI-supported robots should help to counteract the shortage of skilled workers and labour. "The shortage of truck drivers, warehouse workers and dockworkers is a critical factor in global supply chain management," says Marina Bill, President of the International Federation of Robotics. "Robot manufacturers combine their hardware with intelligent software to meet the specific automation needs of the warehousing and logistics industry. Robots equipped with AI open up a huge range of new possibilities for this sector.