Not an ultimate list of digital business solutions for supply chain
April 14, 2021
It takes skill to identify potential benefit sources of modern technologies in different industrial contexts. You must know both the industry you are operating in and the technologies you want to use. In a recent interview about data and technology opportunities in industrial supply chains’ digitalization, I was asked to name the three most beneficial application areas for data relating technologies in manufacturing supply chains. As a rule, you should start with the business target rather than technology. I was a bit reluctant, therefore, to pick any. There are many possibilities. However, I eventually came up with three examples, rather than “ultimate list” of must-dos.
It is my opinion that planning is the core of any operation. Regardless of what you are planning, you need the data and tools to handle it. With effective data handling, it is possible to expand the sphere of our planning activities. For example, the optimum can be searched from a wider group of plausible solutions or wider planning windows in the planning horizon can be used. It may also be possible to capture some very quick or very slow phenomena that are difficult for planners to digest. The continuum from strategic planning to final production plans becomes easier to visualize and manipulate. We can also repeat the planning activities more often, when situations change. In other words, we can move towards continuous planning.
Planning space can then be extended further, by bringing in data from additional external sources. It is possible to collaborate with your customers to get a better understanding of future demand, by integrating directly into their processes using modern integration tools. These data can then be used to understand both single customer’s behavior and market trends in general. The first step might be to use your existing eCommerce platform to provide the data. You could also utilize different external data sets that describe economic trends. it is usually not about getting the data but how to use it to produce meaningful information. It should be both accurate from the data handling point of view, as well as transparent – and understandable – from the planners’ point of view.
It is becoming standard practice to use machine learning or “soft sensors” in production to measure product quality parameters online, instead of sample taking methods that often require laboratory measurements. For those who think of getting rid of these labs: Laboratory results are, of course, still needed to calibrate the models. In any case, the result may be that with the use of, sometimes thousands, signals from the production process, we can generate a real time view of what is going on in the production process. This information can often be used to control the process much more closely than ever before. Quality parameters are easier to keep in their tolerances and production economies can be developed further. This is what production professionals have always done: reduce variance and drive closer to limits – safely.
Once all the internal information has been utilized to the full extent, we can start to think about expanding the data space. If my product is someone else’s raw material, I might be able to connect to their production statistics and see how my product is behaving. You could even guide your customer to get better results or take the knowledge back to tweak your own tolerances to produce a product that works even better in customers’ processes. This information is also important for your R&D. Your raw material may also have characteristics that you should consider, when adjusting your production process. In this example, raw material related data has been relayed in the ecosystem in a collaborative manner over the industrial value chain. Sharing information beyond the commercial transactions may increase the overall value creation significantly.
Humankind has been successful, at least partly, because we are clever enough to share tasks in a way that experts work on what they know best. In industrial practice, this means, for example, that plant operators bring in external parties to maintain their assets collaboratively. You may now need surprisingly many professionals to carry out a relatively mundane service case. This could involve changing a control valve or pump or large, planned, maintenance breaks. Some of the information needed for each case is relevant to all participants, such as the timing of different tasks as well as location and position information. Some of the information, however, must be restricted. This includes commercial terms between the parties. The performance history of certain machinery and equipment may be very important for the two parties but there may be strict agreements on how the data can be utilized or shared.
Another, very important, aspect of data visibility and use with operating manufacturing assets is safety and responsibility for its maintenance. Sophisticated equipment in production units can communicate with the outside world. However, in many cases, we should not allow them to do so. It must be done with strict and tested protocols. Even if we want the data to be transparent, it is ultimately the operator who is responsible for the safety of the people and assets within the operation.
Modern layered data handling technologies allow us to collaborate and share information, as appropriate. However, the visibility can be restricted when necessary and, ultimately, maintain safety. This, however, calls for a delicate design and thorough understanding of both the industrial context and technologies used.
In October 2020, Midagon released the white paper How to drive business value from digital in process…
Far too often driving analytics in companies comes down to single analytics rehearsals conducted by…