- PSI Polska engineers are working on a technology using artificial intelligence.
- The project called Warehouse Intelligence received $3 million in funding from NCBIR and one of its stages was piloted in LPP.
- The use of artificial intelligence has shortened the length of picking paths in the LPP by up to 30%. It is estimated that more than 1/3 of warehouse logistics costs come from order picking.
- The project was co-funded by the National Research and Development Center NCBIR.
Warehouses in Poland
There are almost 24 million m² in Poland. Warehouse and industrial space. As investments soar, pressure on fuel, material, and labor costs is forcing manufacturing and commercial businesses to look for ways to optimize. The logistics industry promises a lot, especially after the automation and robotization of warehouses. The “Logistics in Poland” report shows that 75% of companies see the greatest opportunities for logistics development in this area. Until now, however, applied technologies based on artificial intelligence had a very narrow application. This could change thanks to Polish engineers and scientists. They are working on a technology that has a chance to revolutionize warehouse management: artificial intelligence will continuously offer optimal solutions to people. The first tests show that it does this very effectively.
A challenge game
PSI Polska and a group of scientists from Poznań and Wrocław Universities of Technology have started working on a project using artificial intelligence in warehouses – Warehouse Intelligence. Although not new technologies, their use in logistics is still in its infancy.
– There are technologies on the market that support employees in a very selective way, mainly in the area of the collection process. No one has previously developed a technology based on artificial intelligence, which would be able to optimize all warehouse processes, addressing the problem as a whole – says Jerzy Danisz, Head of the WMS Competence Center at PSI Polska.
– Our idea is to develop artificial intelligence algorithms whose task is to manage the warehouse in such a way as to achieve optimal efficiency of individual processes. The ML model (i.e. machine learning / machine learning algorithm) was given the task and needed to find the best solution. If he succeeded in optimizing a given process, he won. Otherwise, he had to keep trying until he succeeded. In this way, through trial and error, the algorithm reaches the optimal solution, and the warehouse simulation (digital twin) allows rapid and practically free analysis of hundreds of thousands of possible warehouse operating scenarios – explains Jerzy Danisz.
The system optimizes picking
The first results obtained are very promising. It turned out that the use of artificial intelligence reduced the length of picking paths in LPP by up to 30%. It is estimated that more than 1/3 of warehouse logistics costs come from order picking. This process is expensive, especially in high-volume order spaces, such as e-commerce.
-Items for individual orders are collected from storage, then packed and prepared for shipment. Picking and packaging costs are directly dependent on the SKU, i.e. inventory units, and in our case these are significant numbers. Therefore, optimization in this area is of great importance to us. This has a direct impact on the efficiency of the warehouse and the efficiency of order processing for our customers – describes Sebastian Sołtys, Logistics Manager at LPP.
NCBIR saw the potential
The use of machine learning mechanisms with reinforcement proved to be such an innovative approach that the National Center for Research and Development NCBIR awarded funds to fund the project. PSI Polska received a research and development grant in the amount of almost PLN 3 million.
Warehouse Intelligence can be integrated with any WMS class system and any commercial and production company can utilize its benefits. The technology will be available in two versions: automatic (artificial intelligence algorithms – Warehouse Intelligence) and as an analyzer with human support, called PEAR. This term covers a logistics process analyzer that allows you to perform simulations and observe them on a three-dimensional model with its current KPIs. LPP will be able to benefit from this solution in the near future.
Accenture experts predict that by 2035, artificial intelligence will increase logistics efficiency by more than 40%.