Logistics goes AI

The future of AI research

What research fields is Fraunhofer IML pursuing on artificial intelligence in logistics?

Fraunhofer IML is a leader in research on the optimization and digitalization of logistics and production processes. Artificial intelligence is one of the technologies that can develop the greatest potential here: automating processes, using resources more efficiently and increasing adaptability to market changes. However, the researchers at the institute do not see AI only as a technological support, but as a key technology for the development of new innovative solutions that can be sustainably implemented in the real world.

 

Technical progress in artificial intelligence is currently developing rapidly. Logistics provides the necessary context for AI research: real-world applications illustrate the economic and social benefits of artificial intelligence. Logistics research drives technological innovation, particularly when it comes to increasing the economy's competitiveness. 

 

The collaboration with the Lamarr Institute for Machine Learning and Artificial Intelligence – one of five university AI competence centers nationwide that receive long-term funding as part of the German government's AI strategy – gives researchers at Fraunhofer IML the opportunity to help shape the new generation of artificial intelligence at the highest level right from the start. The cutting-edge technological expertise gained from the collaboration with the Lamarr Institute is being put to good use in the institute's research and industry projects.

> Lamarr Institute for Machine Learning and Artificial Intelligence

Big data and AI research in data science and analytics.
© Chaosamran_Studio - stock.adobe.com

What is the significance of AI-supported robotics for science and the economy?

Fraunhofer IML is one of the pioneers of AI-supported robotics: its researchers have been developing autonomous systems that can be used to make workflows in logistics and production more efficient for many years. These include automated guided vehicles (AGVs) as well as mobile transport robots. With the agile and fast transport robot evoBOT®, an innovative platform with an intelligent automatic balancing system, and the highly dynamic outdoor robot O3dyn – both developed in the large-scale research project Silicon Economy – the researchers have set new standards in the section of AI-based robotics.

 

Researchers are opening up new dimensions for the development of autonomous robot systems for logistics, primarily with simulation-based AI. These simulations serve as immersive learning environments where swarms of vehicles can be effectively trained. Individual vehicles, such as mobile transport robots, are represented as avatars, incorporating all physical and control capabilities within a simulated reality. The behavior of the vehicles can be virtually observed and understood in complex scenarios. The final control code can then be transferred 1:1 to the real vehicles without any transformation or adaptation steps. The researchers have created simulation models of the real evoBOT® and O3dyn, which have been published as open source. They enable the realistic evaluation of new designs, sensors and algorithms in virtual reality.

 

Fraunhofer IML sees great potential for automatic planning in AI-supported simulations in conjunction with digital twins – a potentially decisive lever for reducing logistics costs that will strengthen the competitiveness of logistics in Germany as a place of business in the long term.

What role do humans play in AI-supported or AI-based logistics?

Hand pointing to digital interface with AI research technology icons.
© Pereslavtseva - stock.adobe.com

For many years, Fraunhofer IML has been researching the collaboration between humans and machines in an increasingly networked economy. The concept of Social Networked Industry, developed in a Fraunhofer IML flagship project, describes the development of a coexistence of humans and technology towards a cooperation that takes place in digital social networks and in a networked industry or is driven by it. The human-technology interaction approach is becoming more important as AI solutions increase. AI is becoming an active partner for humans. In this context, logistics research is intensively addressing the question of what room for maneuver humans still have or must have when machines in AI-supported logistics not only provide information but can also actively control, plan and negotiate.

The researchers at the institute see it as their task to enable a framework for the cooperation between humans and AI according to the standards and norms applicable in Germany and Europe, but also to implement it technologically. Humans should act as conductors of complex systems that remain comprehensible and understandable, thus promoting trust and acceptance in the working world. A value-based application of AI should also ensure transparency, ethical standards and data protection. 

> More information in the Fraunhofer IML's  magazine “discover LOGISTICS” 

AI research topics across the institute

Numerous current projects at Fraunhofer IML illustrate the potential for AI in logistics.

 

AI for the production of tomorrow

From traditional manufacturing to the data-based factory of the future: The Data Factory.NRW aims to make production future-proof with artificial intelligence. In the research project, Fraunhofer IML is responsible for the area of “Data-driven Logistics”, one of four transformation areas. Specifically, it is about designing artificial intelligence in inbound logistics and supply chains, based on current AI processes.

A quantum leap in cargo logistics

Together with the logistics service provider Dachser, researchers at Fraunhofer IML have developed and implemented the digital twin of a transloading warehouse in the @ILO project – a quantum leap in general cargo logistics. Special AI-based algorithms in the @ILO software identify and localize the packages. The project was awarded the German Logistics Prize.

 

Infrastructure for AI-based logistics

With the Silicon Economy initiative and the corresponding large-scale research project (2020-2024), the Fraunhofer IML has created important technical foundations – an operating system – for a logistics system in which AI systems play a key role. The services, technologies and applications developed in the research project create the basis for the economic use of artificial intelligence in logistics and enable logistics to develop distributed and networked platforms. 

(More than a) Trend-setting technology: Computer vision

Computer vision, a branch of artificial intelligence, is a promising approach to digitalization in logistics. The use of smart cameras in conjunction with artificial intelligence serves to increase process transparency and optimize process flows, thus saving costs and reducing emissions. Experts predict that the technology will become established in logistics within the next five years.

 

SKALA connects AI and blockchain

In the SKALA project, the Fraunhofer IML is investigating how artificial intelligence (AI) and blockchain technology can work together. New AI and blockchain components, including software modules, AI models and smart contracts, make it possible to connect processes in production, logistics and supply chain management across Germany and Europe and to unlock the potential of digital transformation.

AI projects from different fields of activity

In numerous research projects, the researchers at Fraunhofer IML develop solutions with exemplary and model character for a wide range of logistics areas.

AI in hospitals

The 5G-compatible sensor box developed in a research project uses innovative localization algorithms to enable robots to move around in highly dynamic environments such as hospitals and to perform a variety of tasks, for example, to take over the transport of medical products.

Optimization of inland waterway transport

The platform “Oktopus” was developed for controlling the transport of raw materials on inland waterways provides the basis for consolidating inland waterway transport between different players. With the help of AI-supported methods, an ETA forecast and a prediction of the port filling level were developed on the basis of a wide range of data sources.

Efficient on-demand transportation options

As part of the KI4autoBUS research project, Fraunhofer IML has been working on the further development of autonomous shuttles as part of a holistic mobility offering in rural areas. AI-controlled software now creates an efficient on-demand mobility offering. 

The pallet of the future

In the Pal2Rec project, Fraunhofer IML researched the extent to which movement data from (Euro) pallets can be recognized and interpreted using sensors and artificial intelligence.

 

ML Toolbox reduces development costs

ML Toolbox supports the development of ML software: a software kit bundles the most important open source software components for the use of various image-based ML applications and provides them with a uniform interface. Numerous research projects in the AI section are already based on the ML Toolbox because it significantly reduces development costs.

Contact

Anike Murrenhoff M. Sc.

Contact Press / Media

Anike Murrenhoff M. Sc.

Deputy Head of Department

Phone +49 231 9743-202

Fax +49 231 9743-162

Benedikt Mättig

Contact Press / Media

Dr.-Ing. Benedikt Mättig

Member of the Scientific Board

Phone +49 231 9743-270