Features and innovations

Data Science Project: Technilog Technik und Logistik GmbH relies on the "Warehouse Healing" product innovation

As part of a pilot project, Technilog decided to carry out a proof-of-value of the "Warehouse Healing" strategy in February 2021. In the process, the data science experts from S&P validated the high benefit and potential in the warehouse using concrete customer data. "Warehouse Healing" is one of a total of 3 strategies for optimised decision-making in the warehouse.

Exterior view of the Obeta logistics center

Our long-standing customer Oskar Böttcher GmbH & Co. KG and its logistics subsidiary Technilog pursue the approach of continuously increasing logistics performance and thus the level of service for their own customers. To achieve this, Technilog not only relies on forward-looking automation technologies and the continuous expansion of its warehouse capacities, but also on the use of data science and AI. With its fleet of vehicles and their optimally coordinated deployment, the company makes daily deliveries throughout Germany. "We have been dealing with the use of data in logistics for a long time and it is becoming apparent time and again what opportunities this offers us in the warehouse. Together with the experts from S&P, we have now taken the step of carrying out a potential analysis. With the help of existing data and intelligent models, we want to uncover unused potential in our warehouse. One thing is already clear to me: an extremely exciting and forward-looking project," says Nico Schubert, project manager at Technilog Technik und Logistik GmbH.

Creating real added value from concrete data
Under the motto: "Data is the new oil", the existing data from Technilog's logistics form the raw material for the proof-of-value. Before the data can be used and real added value can be generated from it, the relevant data such as warehouse stocks, topology, item information, shopping baskets and movement data are identified in Technilog's warehouse. The data basis is provided by the warehouse management software SuPCIS-L8 from S&P. The data is then visualised and interpreted. Algorithms were used to determine stock transfer and exchange suggestions and, by interpreting the results, the savings in walking time specific to the use case were determined. "The storybook answer now would probably be: collect data, analyse it, model it, put storage and stock transfer suggestions into action and thus create real benefits. However, the use of artificial intelligence also brings with it its requirements for both sides, which had to be tackled together and implemented in the best possible way," continues Timofej Woyzichovski (data science expert, S&P Computersysteme GmbH).

The new strategy "Warehouse Healing" is designed to defragment the warehouse and, through the intelligent analysis of movement data and baskets of goods, to reduce the travel times for man and machine through better positioning of items. Starting from a strong warehouse fragmentation, the two project partners focused on the analysis, evaluation and pattern identification of the prevailing situation: this data is the basis for the next phase in which the model was created and trained.

The next step was to define the score of the manual pallet warehouse as a storage area. In order to find stock transfers at this point, after which the score is better than before, algorithms are used that include the "experience" from the order history of Technilog's customers. The results of the algorithms provide the basis for simulating stock transfers and defragmenting the warehouse beyond normal levels - this process was done without any impact on Technilog's ongoing business processes. The beauty of this is that the automatic training experiments in the background continuously improve the results.

From artificial intelligence to real use in the warehouse
The "Warehouse Healing" strategy enables Technilog to draw the full potential from predefined use cases: from data collection to use in intralogistics. In the process, it is important for the data science experts at S&P to jointly evaluate the results and discuss the expected benefits. At Technilog, the results from the proof-of-value showed that the application of warehouse healing in the storage area "manual pallet warehouse" already reduces the retrieval paths by approx. 24 % after a few hundred stock transfers. The logical step - the use of the warehouse healing strategy in daily operations - was decided immediately after the presentation of the significant results. "With the Proof-of-Value, Technilog has taken an important step towards uncovering hidden potential in the warehouse. We are pleased about the joint success, the trust placed in us and the valuable input from our customer Technilog in the further development of our Data Science products," Rémy El Abd, Managing Director at S&P Computersysteme GmbH, adds in conclusion.

Would you like to learn more about the "Warehouse Healing" strategy or further information about the project?Feel free to contact us!

Contact us and talk to one of our experts on the subject.