Big Data from Sensors

Everyone is talking about AI and it's a big topic for the future, especially for industry. How far along is the development of AI in your business field, i.e. intralogistics?

Björn DunkelIf we look at supply chain management in principle, there are already a lot of areas where we can talk about AI. For example, in route planning, where the route is suddenly changed due to traffic jams because a supposedly longer route now means a shorter travel time. At GIB, however, we are concerned with the internal supply chain, i.e. everything that relates to intralogistics. Here, too, there are already approaches to how, for example, sales forecasts can make better and more reliable statements through the use of artificial intelligence. With the help of an artificial neural network, software can now learn from mistakes so that it achieves better and better results over time. We are certainly still a long way from achieving this today, but we are already well on the way.

In which subarea of intralogistics can you imagine the use of AI or perhaps even already work with it?

DarkAt GIB, we offer software specialized in the supply chain for sales planning, production planning, collaboration with suppliers, material requirements planning, inventory management and operational planning by refining the SAP standard. Especially in quality assurance at the end of production processes, there are already approaches for the incorporation of AI. Namely, where image recognition is used to distinguish between good and bad parts. By means of an automated exclusion process, only good parts reach the further processing stage. This is a big issue because this exclusion process is still done manually in many factories today. The coupling of 3D image recognition and artificial neural networks, for example, offers massive optimization possibilities in quality assurance here. Thus, NIO parts can be inspected directly in the production process without human intervention and automatically assigned to the next production step. This eliminates potential sources of error, streamlines the process in terms of time and immediately detects the production of defective parts.

Quality assurance would therefore be an area of application for AI. Where are you already using calculations based on neural networks at GIB?

DarkWe are still in the early stages of using AI in our software. And in fact, this is the first time that we have used a real AI in the environment of our supply chain solution that is capable of determining an exact safety stock through learning via neural networks. Now, however, the system must first prove itself in real use. Only then will we see whether our AI also works in real life. But we are already working on other possible uses for AI. A team of developers is already researching beneficial applications in sales forecasting.

The keyword is big data. Today, gigantic amounts of data are generated through the use of sensors. What happens with it?

DarkBig Data is of course not the goal, but at most a means to an end. In the case of the use of AI, this is a blessing in that an AI only works successfully if it is fed with a great deal of data and information. In the learning process, "a lot helps a lot" really does apply, at least if the formulated question was correct and the algorithms work well. The fact that we are now able not only to collect these gigantic amounts of data, but also to store them and make them available for further use, is the rocket propulsion that really got the use of neural networks going. So what happens to the sensor data that is collected at the production level?

How have you solved the challenge so far?

DarkUntil now, the evaluation of the same was rather close to the machine. That is, the health status of the machine could be read on the system, in the plant. Here, there was a strict separation between OT and IT, i.e. between production and business process control. Better security concepts and technologies have increasingly led to an opening up of production IT, not least because digitalization makes this necessary and the advantages of vertical networking of supply chain processes offer an enormous competitive advantage. At GIB, we have developed a solution that is able to bring aggregated sensor information into the ERP system. This means that we not only digitize business processes horizontally along the value chain, but are also able to digitally map the production level. I am certain that the targeted use of AI will be a decisive success factor, especially in the meaningful use of Big Data.

How do you plan to make better use of big data with the help of AI?

DarkData only makes sense if you put it in context. When I walk through the factory, I want to see what has been planned at the ERP level and be proactively alerted by an end device such as a cell phone that productivity is particularly high or particularly low at the machine I am standing in front of. But this information alone is of no use to me; I have to put it in context. In the mass data that the sensors collect, there is information about parameters that can influence my production, such as the air pressure, the temperature in the hall or the humidity. And if I contextualize that now, I know which jobs were running on this machine at the time. I know what capacity utilization the machine had, what articles were produced on it. And if I notice again and again that quality problems somehow occur at a certain point in time, I can look at the data that occurred at the same time on the store floor level.

What can the user conclude from this?

DarkFrom this, derivations can be made. With slide rule and pencil such derivations and correlations cannot be processed. There are too many variables, which all influence each other. An AI could be the solution here. The AI recognizes patterns and correlations, evaluates the situation in its entirety and provides best possible solutions for the respective addressee, for example for the best production sequence in the given situation. Overall, contextualization is a key point in the optimization of intralogistics. We like to use the term CLUI, which stands for context based, location based and user based information. So our goal is to address the right info correctly, exactly where and when you need it.

So production planning can be optimized through Big Data and AI?

DarkYes, exactly. An incredible amount of data is available via sensors. That means we have a data pool that we can use. AI automatically makes the decision as to which product mix is the optimal one with the temperature, humidity or other parameters forecast for today. This can significantly reduce the production of bad parts. These are intelligences where we are not that far away from because this pairing of know-how is already there.

Is this how OT and IT levels grow together?

DarkYou could say that. In this context, we like to talk about the "Y-way", we bring together Operation Technology and Information Technology in one system. Of course, not all store floor data can be fed into the ERP system. The terabytes of data would explode any ERP system. The trick is to select which data plays a role in the digital mapping of the production level in the ERP; in other words, what is needed to bring the "digital twin" to life. We then pull this data into the mass data store and can use it later for evaluations and interpretations, for example using AI.

Is it conceivable for you that in the future, thanks to AI, a warehouse will send its own orders to suppliers?

DarkYes, absolutely. Our Vendor Managed Inventory (VMI) tool is already heading in the right direction as a digital link to the supplier. You may not believe it, but 90 percent of B2B orders to suppliers today still go through an e-mail. That means that only when an e-mail arrives does the supplier know what the customer actually needs from him. Our vendor-managed inventory solution creates transparency and plannability for the supplier and at the same time relieves the customer in his demand planning. The supplier can view all the information relevant to him at any time and use it to plan, for example, which delivery call-offs are planned and which stocks will soon be depleted. Another positive effect is that the supplier can use it to make his own processes more economical and cost-effective. This in turn often has a positive effect on the terms and conditions with the customer.

So is the reorder point decisive?

DarkSo if we look at our existing VMI solution to see where AI might be useful, then the reorder point would certainly be a good place to start. Referring to our climate example from earlier, the empirical values from the past could be related to the external factors, such as humidity, pressure and temperature. A smart AI evaluates the current situation, recommends an adapted "recipe" for production, and triggers the required delivery notification from the best supplier in good time.

Does the use of AI bring us a little closer to the perfect supply chain?

Dark: That is indeed our goal. If we look at the indicator system of our Supply Chain Excellence solution, SCX, for example, we give the supply chain manager a central indicator that tells him at a glance how good the processes in his supply chain are. This metric is of course based on a number of key performance indicators, which in turn are based on the SAP Data Core, i.e. Big Data. Currently, our supply chain evaluation system is designed by many complex algorithms and as we add more and more intelligence and thus more and more key performance indicators to the consideration, the complexity of the system also increases enormously. We will eventually reach our limits with this classical way of programming. That's why we will focus on using artificial intelligence here in the future.