A new way to reproduce tacit knowledge and its usability in logistics

Our experience from customer projects at CAMELOT Management Consultants has shown that process automation and digitalization have changed the way businesses operate today by boosting speed and improving efficiency. As a result of this, human interaction with computers, machines, even with robots and other cyber-physical systems, has become an integral part of our today’s business life.

However, it is exactly the human behavior that is still irreplaceable. We all know more than we can communicate or write down. Humans have built up an extensive amount of tacit knowledge, which we have not managed to fully translate into a programming language yet.

Have you ever met a machine that is good at telling jokes?

Well, probably not, but that is the actual problem: No one has yet been able to formalize the process of inventing funny jokes. Humans act reflexively as they unconsciously combine and merge thousands of impressions every second.
It is our tacit knowledge that makes us unique in being able to simultaneously process a variety of variables without even recognizing it. And tacit knowledge grows over time, allowing us to become faster and more accurate in making a decision.

Artificial intelligence is changing the game

New technologies are emerging, bringing new buzzwords such as machine learning, artificial intelligence, or neural networks – and precisely these new technologies will be able to fill the gap of translating humans’ tacit knowledge into a programming language.

Artificial intelligence, in the form of neural networks, can be understood as an IT model of cognitive structures that are able to reproduce the information processing of human brains. They are generally self-learning algorithms that can synthesize tacit knowledge themselves: Based on huge amounts of input variables and related historic data that is revolving every day, they can detect patterns that make them think in a way we cannot express.

Furthermore, neural networks can differentiate between those variables that add value to making a decision and other variables that are just noise. Over time, they can improve in order to provide even more accurate responses.

How can it be applied in logistics?

Especially in logistics, frequent staff turnover is very common, as many logistics employees are dissatisfied with advancement opportunities or simply stressed by an unsatisfying work-life balance. Often, logistics employees move internally to positions outside logistics with a better image. That means that tacit knowledge is often lost and optimization potentials cannot be fully exploited, as tacit knowledge must be re-established repeatedly. Therefore, companies should build up tacit knowledge using artificial intelligence, which will allow to conserve and reproduce a continuously growing knowledge base.

Another advantage is that neural networks discover, learn, remember, compare and reproduce data patterns a lot faster than any human being. A neural network can be trained so that it works like an employee who has hundreds of years of experience and a perfect memory – saving time and money.

The list of action fields in logistics is long, and in order to transfer what you have read in this article to what you encounter every day, we want to provide you with a day-to-day example. Let us think about the daily business of a dispatcher in logistics who is able to unconsciously anticipate future behavior of customers due to extensive experience with their attitudes. In this respect, neural networks must be able to learn and mimic the patterns that a dispatcher permanently has in his mind.

Reproducing tacit knowledge by artificial intelligence will, for example:

  • increase predictability of supply chain disruptions
  • increase predictability of transportation demands before the customer even places the order
  • increase predictability of vehicle maintenance
  • increase vehicle capacity utilization based on different variables and not on a static rule for defining freight units
  • decrease picking distances by continuously re-arranging your warehouse layout
  • decrease costs of human failures

What you should do now

As you can see, there are many application areas where artificial intelligence can be used in logistics. However, as a first step, it is important to understand your company’s current degree of maturity with respect to logistics digitalization: The backbone of all digitalization efforts is still the fundamental improvement of any as-is operations. And in doing so, access and availability to related data and information is essential.
Thereafter, you must gain an overview of possible technologies, concepts and developments in logistics digitalization while understanding your company-specific future requirements. In doing so, you will develop an understanding of future action fields in which you would like to dig deeper, recognizing dependencies between those, in order to develop a prioritized road map for implementation. However, keep in mind that future logistics solutions must be interlinked with all supply chain steps to unfold their full effectiveness.

Sounds complex? It certainly is. But nothing ventured, nothing gained!

Here you can find more information on the topic digitalization in logistics and how to start it: the path to leaner and more cost-efficient logistics.

written by
Steven Malbrant
Steven Malbrant is consultant in Logistics Practice at CAMELOT Management Consultants. He is predominantly involved in projects in the chemical and pharmaceutical industries, covering both strategic and operational topics. To outline some of his focus areas, Steven is engaged in supply chain processes, network design, and transportation management.

This post is also available in: German