Processes are often seen as a pearl necklace kind of chain that has a clear start and end point and where the subtasks are run separately from the rest of the business. However, the dependencies between processes and the impact of processes on their environment is a significant issue. Sometimes dependencies may cause unwanted surprises and restrictions and sometimes they offer new kinds of business opportunities that may not have been outlined earlier in the basic processes.

If the processes are controlled without an information system, the effect of different dependencies can be noticed by ad hoc principle. Challenges are solved as they arise, and they may not even be perceived as big. The situation gets complicated quickly if the process is digitized poorly, and the information system intended to support the process consolidates practices into inflexibility, and does not take into account other processes and external dependencies.

How to discover a model where everyone wins?

The solution may cause sub-optimization and therefore one process gets intensified, and the others suffer. In the worst case all of them get disadvantaged. Digitalized process becomes too stiff and at the same time other processes become more difficult.

The risk becomes particularly high when the work tasks are automatized. The algorithm or artificial intelligence that replace man-work, cannot notice the environment outside the process and create an overall picture in the way that a human can.

However, digitalization can also go well. In the best case, the interdependencies and links between processes and their environment are already clear to all of the operators and the overall impact of process changes are viewed in the early stage. Therefore, it is possible to reach the win-win mindset.

The consideration is not only the efficiency of one process, and not only the internal development of the company, but also what are the effects on the company’s customers and their processes. It is under discussion how to support green transition by process improvement in personal and in client’s action.

Case SA-TU Logistics

These questions were discussed also in SA-TU Logistics Oy, when they participated to Finland’s first artificial intelligence accelerator of clean industry implemented in the AI-TIE project in autumn 2021. The goal was to find ways to develop operations with the help of artificial intelligence. At the first stage we created ideas and analyzed several potential use cases. At the final stage, after many conversations we picked one of them: customs clearance completion time forecast. The choice was clear because the use case offered ways to optimize several processes and count the emissions caused by the operation.

SA-TU Logistics is a domestic company offering logistics services. Since December 2021, the company has been part of the Customs Support Group of more than a thousand customs professionals operating in eleven European countries. It has long been the largest privately owned producer of customs declarations in Finland with approximately 150,000 annual customs declarations. Customs clearance is a demanding specialist work which is performed with the help of IT solutions and in the future also even more with the support of artificial intelligence solutions.

It is very important for the company’s customers to get a reliable forecast of the time required for customs clearance, so that the order for onward transportation of the customs-cleared products from the customs warehouse can be precisely scheduled. In this way, unnecessary driving kilometers and unnecessary waiting, and the resulting emissions and costs, can be avoided.

The objective for SA-TU Logistics is to find ways to optimizing the use of man-hours and boosting the time efficiency in processes. The possible solutions also have a direct impact to the surrounding processes. Possible solutions also have a direct impact on the surrounding processes.

At the moment the transport companies do not know when the declared cargos are ready for pickup. Therefore, the truck may arrive too early just to make another run before pickup. The truck may also be waiting without a task for customs clearance to be completed. This results in unnecessary staff costs for transport companies, because the drivers also have to be paid for the waiting time. There is also unnecessary fuel expenses. Because of the huge volumes, the unnecessary emissions have a huge impact.

Approppriate and viable object for development

The solution is to develop a new feature for the customs clearance order portal currently used by SA-TU Logistics’ customers, a forecast for the duration of the customs clearance assignment. With the prediction a client can follow the duration of customs clearance and the estimated time for completion and, for example, organize the transport capacity according to it. At the same time, the system produces information for SA-TU Logistics’ internal use and enables even more accurate planning of human resources and time use.

Feasible and suitably small development targets were searched for in the artificial intelligence accelerator. Those that are easy for companies to get started with, even if there is no previous AI expertise. The customs completion time forecast produced with the help of artificial intelligence turned out to be a suitable challenge. The choice was also led by the fact that SA-TU Logistics had enough reliable data on their use.

We chose this development target (forecast of the completion of the customs clearance assignment) because its data is relatively easily available and we know that it can be implemented not only as an artificial intelligence solution but also, if necessary, with traditional methods, so in no case will the work be wasted.

IT Manager Jyrki Messo, SA-TU Logistics Oy

What’s next?

The approach used at the accelerator proved to be effective for SA-TU Logistics. From this and other business cases of the AI-TIE project, more experiences and lessons learned will be collected in the future to share. They will be summarized as part of the online course to be published in autumn 2022.

At best, the planned course guides the company to choose the most promising objects and use cases to be developed with the help of artificial intelligence. From here, it is easy for companies to continue, for example, to the Proof-of-Concept (PoC) phase, to the more detailed planning of the artificial intelligence project and to the selection of a suitable development partner. Let’s bring artificial intelligence to work together!

Translated from the Finnish by Sari Benford and Petteri Saloranta.

Photo: www.shutterstock.com