Artificial intelligence in logistics: status quo & potentials

Artificial intelligence (AI) now offers a wide range of applications and opens up many new possibilities for logistics companies. However, it has by no means fully arrived in the industry yet. Highlights from two forums at transport logistic 2019 shed light on the use of AI in logistics and its potential for logistics, supply chain management and the airfreight industry.

The use of artificial intelligence in logistics

Logistics, with its widespread networks, is an ideal application area for artificial intelligence. The intelligent analysis of data, for example, helps to forecast future production and transport volumes, enabling to use resources more efficiently. Today, tasks are therefore increasingly delegated to self-learning digital systems.

What is artificial intelligence?

According to the Encyclopedia of Neuroscience, artificial intelligence (AI) is an area of computer science that deals with the exploration of mechanisms of intelligent human behavior (intelligence). This is done by simulation using artificial artifacts, usually computer programs on a computing machine.

AI in logistics: the status quo

AI has enormous potential, but has not yet arrived in logistics, particularly not in the airfreight industry, as a study by the German magazine Logistik heute und INFORM demonstrates. The survey among a total of 123 employees and managers from various logistical disciplines delivered the following insights on the current use of AI in logistics:

  • Around 90 percent of the respondents expect artificial intelligence to improve their market position.
  • However, only 26 percent of the companies surveyed are already actively using AI in their logistics processes.
  • The reason: more than every second (54 percent) lacks the know-how. Only around 12 percent of the respondents already feel well informed about the topic.
  • In addition to this knowledge gap, other obstacles to the implementation include high costs (46 percent), inadequate IT infrastructure (44 percent) and lack of time resources (38 percent).
  • About 60 percent of the respondents stated that the topic had hardly penetrated logistics so far.

Our conclusion: Today, only few companies already rely on artificial intelligence in logistics. The main reason seems to be a lack of knowledge. This already begins with the definition: in the past, we used to give the computer data and—with given algorithm—the computer calculated. Today, AI determines the path from data input to data output itself. Robotics, speech recognition and, as a subarea, machine learning also function on this basis.

AI in logistics: Where are the greatest potentials?

Data and algorithms allow forecasts to be made under the influence of a wide variety of parameters. But what can AI offer logistics? According to the survey, logistics companies see the greatest potential in the following areas:

  • demand forecast and sales planning (62 percent)
  • production optimization (51 percent)
  • transport optimization (50 percent).

Various practical examples show what AI can achieve in transport and logistics, for example:

  • organize refueling tours with filling levels
  • design electric vehicles considering temperature, topography and traffic
  • use drones to send images to containers for damage testing
  • differentiate dangerous goods by means of image recognition (even today, neural networks can recognize 27 different symbols).

The provision of short-term information via AI also saves time and money. This includes for example:

  • traffic jams
  • sudden changes in the weather
  • waiting times at the customer.

The forum participants at transport logistic 2019 see further advantages of AI in event-based and dynamic route planning.

Finding areas of application for AI in logistics

But before implementation, it is important in the first step to be able to identify application areas for AI at all. Because long before the algorithm, the first thing to do is to integrate data. DB Schenker currently has 20 smaller and larger use cases, some in research. The focus is on

  • dynamic offers and prices,
  • demand and forecast as well as
  • capacity planning and
  • autonomous vehicles.

Artificial intelligence can for instance also help to analyze the booking behavior and recognize when a customer is about to leave or pack goods into containers efficiently and in a space-saving manner, like a 3-D Tetris.

Read our article “AI application areas in logistics“ to learn more about where else in logistics AI can be applied and thus help to counteract skills shortage.

The added value of AI: the self-learning supply chain

To this day, people are talking on the phone, sending e-mails and faxing. Prices are negotiated orally, while static data provides orientation. Given the multitude of combinations of time, paths and resources, however, no one can find the mathematical optimum. The aim and added value of AI is therefore in particular the self-learning supply chain.

The advantage: It avoids that there is no truck available for transporting products coming out of production, which would mean delays in delivery and that the customer misses company goals. However, this can only work without a silo mindset and rigid function boundaries.

The use of AI in the airfreight industry

Thanks to globalization and worldwide Internet trade, the demand for fast airfreight is also rising. According to a study by the German Aerospace Center (DLR), its volume will have more than quintupled between the turn of the millennium and 2030. The problem: currently, most of the processes involved in loading freight on the ground still are carried out manually.

One of the forums at transport logistic 2019 also dealt with this important topic. „Artificial Intelligence: Next Level Air Cargo?“ demonstrated the interest of the airfreight industry accustomed to success. The problem of the airfreight industry primarily is the increasing customer demands. The passenger segment with its online portals and the ticket service provider Etix on all terminals sets the benchmark here.

For the complex value chain, the forum's discussants see advantages in a cloud or at least in standards for data and processes. This is what the big players in the International Air Transport Association (IATA) work on.

The usual mindset is the biggest challenge. The industry is still reluctant to share its data. But it must, so the consensus: “There is a killer out there,” warns Thorsten Friedrich, who introduced e-billing at Lufthansa, strikingly. His view into the future: “The first to control data at Amazon level will shake the airfreight industry.”

You are looking for more information about the forum “Artificial Intelligence: Next Level Air Cargo”? Here can be found a video of the entire panel discussion.

Read our article Developments & Logistics Concepts of the Future to find out about the other trending topics in the logistics industry.