Fraunhofer Institute for Material Flow and Logistics IML
Fraunhofer Institute for Material Flow and Logistics IML

Digital transformation in logistics with AI technologies

AI in logistics

Content

Is artificial intelligence in logistics ready for use?

Artificial intelligence optimizes route planning and capacity utilization, transforms simulations into learning environments for swarms of vehicles, minimizes machine failures through predictive maintenance and facilitates smart finance that enables new business models. Consequently, artificial intelligence (AI) has sparked a technological paradigm shift in logistics – the potential for AI processes and methods in the industry is virtually limitless.

Symbol image for artificial intelligence in honeycomb-like, blue glowing patterns
© putilov_denis - stock.adobe.com

Fraunhofer IML is one of the pioneers of AI-based logistics. Its researchers are developing innovative solutions for industry based on the latest scientific findings. Thanks to big data, increasing computing power, new training methods, open source frameworks and a growing digital transformation, logistics is entering a new dimension with AI.

AI systems are capable of optimizing complex logistical processes and taking on both recurring and creative tasks that were previously solved solely by human knowledge and experience or by human intuition. This opens up completely new possibilities for companies to optimize, automate and autonomize their processes. “Logistics is characterized by the collaboration of many different actors in complex value chains and networks. Often, conflicting goals have to be reconciled. This is where the new methods and procedures of artificial intelligence come into play: they are able to analyze and control the complex relationships in the logistics environment faster than humans can today,” explains Anike Murrenhoff, AI coordinator at Fraunhofer IML. “AI is thus becoming a driver for tomorrow's logistics systems and has the potential to fundamentally change the logistics industry.” At the same time, artificial intelligence is also seen as key to the future of medium-sized companies.

To work effectively, AI applications require high-quality data. In logistics, such data is already being generated and collected from a variety of sources, from sensors to robots. This data provides the basis for a wide range of methods and processes of artificial intelligence, such as machine learning, computer vision and generative AI.

Profile photo Prof. Dr. Alice Kirchheim

“The use of artificial intelligence in the industry requires close cooperation between research and companies. One success factor lies in understanding where AI can specifically help companies and increase the potential of innovation.”

- Prof. Dr. Alice Kirchheim, Institute Director

Research and practice: AI for logistics

Predictive analytics in production and logistics

Ressource management

Quality control

Decision support systems

Knowledge management

Process optimization with AI

Robotics

Intelligent automation

Smart sensors

 

What can AI do better, faster and more effectively than humans in logistics?

The aim of artificial intelligence is to replicate and imitate the intelligent behavior of humans. In its “Application model of machine learning in logistics”, the researchers of Fraunhofer IML distinguish between three different capabilities of artificial intelligence: sensing (detection), thinking (analysis) and acting (acting, planning and deciding). Accordingly, artificial intelligence can play to its strengths in the logistics areas of procurement and purchasing, production, and sales and distribution. In each of these sections, there are physical activities for the transportation, handling and storage of goods, as well as the associated planning and control tasks.

Icon: Sensors monitor logistics objects – AI in logistics

Sensing

The field of sensing applications is based on the use of sensors that collect information about the environment – whether it be motion, temperature or image sensors built into cameras.

Logistics use cases for sensing with AI:

  • Monitoring of stocks
  • Monitoring of supply chains
  • Detection of logistics objects
  • Recognition of hazard labels
  • Recognition of individual workpieces and transport aids
Machine learning icon analyzes data – AI in logistics

Thinking

In thinking use cases, machine learning algorithms are being used to make decisions as well as to analyze the generated data, for example, in use cases related to sensing. 

Logistics use cases for thinking with AI:

  • ETA predictions
  • Personalized break recommendations
  • Load formation and volume measurement
  • Predictive maintenance
  • Detection of anomalies/damage
  • Assistance in maintenance
Icon for robot arms perform gripping processes – AI in logistics

Acting

Logistics use cases of acting include the use of robotics and automation to perform physical tasks.

Logistics use cases for acting with AI:

  • Voice control through pick-by-voice systems in order picking
  • Control of autonomous vehicle swarms or automated guided vehicles (AGV)
  • Gripping processes of robot arms
  • Routing of autonomous vehicles through rack aisles

Further reading: Artificial Intelligence in logistics

Fundamentals from research

The Whitepaper “Künstliche Intelligenz in der Logistik” explains the basics of intelligent processes and methods, as well as current developments and future fields of application for AI in logistics.

Download

AI in practice

“Industrial AI – Artificial Intelligence on its Way from Hype to Practice” is a key topic of the magazine "Discover logistics": with insights into new AI-based developments and practical tests, from airport tarmacs to hospital logistics.

Download

Use-Cases from the industry

In the White paper “AI in logistics” of the technology platform Alliance for Logistics Innovation through Collaboration in Europe (ALICE), companies present applications based on artificial intelligence.

Whitepaper

What does the rapid development of AI mean for the logistics sector?

Whether it's intralogistics, production or transport: artificial intelligence can be used in all areas of logistics and is now accessible to companies of all sizes. The costs for developing and using AI solutions have decreased in recent years, and the acceptance of AI systems is also increasing among employees. Artificial intelligence has therefore become a way for companies to prepare themselves for the future.

Abstract image of a glowing brain – AI in logistics
© improvee design - stock.adobe.com

Most companies view AI as a strategic tool for improving their logistics processes and are pursuing an exploratory approach. They are keen to get an overview of the possibilities offered by modern AI and how it can improve their processes. This can lead to new ideas and innovations that may exceed initial expectations. At the same time, companies can also approach the topic by considering a specific challenge or issue that needs to be resolved. The problem-centered approach allows for faster, more tangible results, and the success of the AI application can immediately be measured. 

Until now, companies have had limited opportunities to cultivate their own expertise and resources for developing or integrating AI solutions independently. This can be partially attributed to the rapid evolution of the topic and the scarcity of specialists. In light of this, researchers at Fraunhofer IML are assisting companies with targeted offerings and projects for the application of artificial intelligence.

Our research and services

© Improvee design - stock.adobe.com

Our offer for companies

Fraunhofer IML supports companies in making existing production, logistics processes and value creation networks future-proof with artificial intelligence – to increase their efficiency and secure their competitive advantage. The services range from the development of customized AI strategies to specific, ready-to-use solutions. In doing so, companies benefit from the latest scientific findings and cutting-edge AI technologies and solutions. 

MORE ABOUT OUR OFFERS
 
 

© Chaosamran_Studio - stock.adobe.com

Our research on artificial intelligence

Fraunhofer IML is making artificial intelligence a reality for logistics. Together with the Lamarr-Institute for Machine Learning and Artificial Intelligence, researchers are advancing new approaches in the development of AI. These include AI-based robotics and AI-supported simulations, which serve as learning environments for robot swarms. Another focus is on computer vision, the collection and analysis of visual data, which is used, among other things, in intra-, yard- or transport logistics.

MORE ABOUT CURRENT RESEARCH PROJECTS AND RESULTS

FAQ about artificial intelligence in logistics

Expand all Close all

  • Artificial intelligence (AI) refers to the ability of machines or computers to perform tasks that normally require human intelligence.

    There are different approaches to developing AI. These range from simple algorithms to more complex systems such as machine learning (ML) or neural networks, which enable machines to learn and improve from experience – in other words, data – and improve themselves.

    Unlike conventional algorithms, no solution path is modeled for self-learning systems. Logistics is an area in which enormous amounts of data are generated and processed. AI can analyze this data efficiently and recognize patterns. This means that forecasts for demand or route planning, for example, can be far more accurate.

  • Machine learning (ML) is a subfield of artificial intelligence and is of central importance for AI applications. The theoretical methods were developed decades ago. But it is only with the increasing volume of data and the diversity and quality of data, combined with the increase in computer performance, that ML methods have arrived in practice.

    ML programs differ from conventional software in their fundamental adaptability to changing circumstances. While changes in requirements in conventional software require adaptation by software developers, ML programs are based on the idea of human learning and develop independently on the basis of new system data.

    > More about the advantages of machine learning in the optimization of supply chain management

  • The logistics industry is predestined for the use of AI because it manages complex processes with many variables, where artificial intelligence can play to its strengths. For example, logistics processes often require quick decisions, such as in route planning or inventory management. AI-supported systems can process real-time data and instantly make decisions to avoid bottlenecks or better distribute resources. 

  • In addition to new, future-oriented areas of application such as simulation-based AI or image recognition and interpretation (computer vision), AI and ML can also be used to replace existing forecasting and optimization methods, thus achieving more accurate results.

    Important areas of application include demand forecasting, sales planning, transport optimization (e.g. through autonomous transport systems) and production optimization. One of the strengths of AI is its ability to predict future developments based on historical data and current trends. AI also makes it possible to automate repetitive and time-consuming tasks, such as administrative staff work related to inventory or processing orders.

  • Artificial intelligence is one of the most important levers for logistics to automate processes in a profound way. AI ensures that processes can run faster, more precisely and autonomously. Human errors can be reduced and resources can be used more efficiently. The use of artificial intelligence ranges from automated storage and picking systems to autonomous vehicles and the predictive maintenance of machines. AI not only increases efficiency, but also the flexibility and adaptability of logistics systems.

More about this topic