Artificial intelligence optimises the packaging process for Kaskipuu windows

Dynamik helps door and window manufacturers achieve significant cost savings by automating the window packaging process.

The volume product of the Kaskipuu is windows, which are manufactured in a wide range of heights, widths and depths. The products are delivered to customers mainly by trucks or full trailers. However, the cost of freight transport is always the same, regardless of the number of Kaskipuu windows that can be transported. “The more windows we can fit per pallet metre, the lower the cost per product, and in the past we have not even used the full height of the load space,” says Mika Salmela, Logistics Procurement Manager.

Optimizing the packaging process started with understanding the problem

Kaskipuu is Finland’s largest family-owned door and window manufacturer, supplying more than 100 000 products this year. Volumes of this size mean that transport costs can be quite high. A preliminary assessment of logistics operations showed that freight is the company’s biggest expense item. “Because of our high volume, we knew that even small savings would have a significant impact on our costs,” says Mika Salmela, Procurement Director, Logistics.

The Kaskipuu team quickly realised that the dispatchers would spend too much time manually figuring out how to pack and place the windows in the hold – which would only slow down the packing process. So the idea of using AI to help with the calculations gradually emerged.

Kaskipuu already knew Dynamik’s employees and knew that they were highly skilled and experienced in AI technology. The company therefore decided to proceed with the project together with Dynamik, who carried out a thorough needs analysis and extensive background research, including numerous visits to the production facilities to clarify the process.

There are many parameters involved in the packaging optimisation process. Firstly, it is necessary to find out how the windows of the order will be placed in the packaging and how these packages will be placed in the boot to optimise the available space. An additional parameter is to ensure that unpacking is easy at customer sites, which can often be challenging to deliver to. By the time the first window in a customer’s order leaves the production line, the person handling the order already knows the best packaging solution for shipping, because the AI has already calculated the most cost-effective way.

“For example, positioning the windows sideways could allow two windows to be stacked on top of each other to take advantage of the full height of the boot. However, calculating this manually would be complicated and almost impossible. Instead, we now just enter all the parameters for the order into the system and the AI does the work for us almost instantly,” says Mika Salmela.

This optimization process can be thought of as a giant game of Tetris: a specific space needs to be fitted with the largest possible amount of different sized rectangular objects possible.

Dynamik’s AI solution has been in active use since spring 2021. Initially, staff were thought to be resistant to using the solution for fear that it would make their work harder and slower – but it was quite the opposite!

According to Mika Salmela, optimising the packaging process has also brought two unexpected additional benefits. The company now uses fewer pallets and less packaging material. The amount of material previously used to pack three windows is now enough for eight windows.

Both Pauli Saarenpää and Mika Salmela praise the work and flexibility of Dynamik’s experts. Such a project presented many challenges, such as parameter changes and incorrect data entered from Kaskipuu’s ERP system, which is currently being fine-tuned. Another challenge was non-parametric factors, such as the unusability of the calculation if a window was missing from the customer’s order for some reason. “These challenges have been overcome quickly and good solutions have been implemented without delay. Dynamik has shown its strength by reacting very quickly to all the necessary changes.”

Mika Salmela estimates that the fine-tuning of the system will probably take the rest of the year. Nevertheless, the benefits are quite obvious, the most significant being the cost savings. The packaging process is now much simpler and more efficient. Salmela estimates an increase in efficiency of at least 30-40% so far. Optimisation of packaging and storage space, and in the future optimisation of transport routes, are part of the planned development process. Artificial intelligence can offer endless possibilities.

Kaskipuu expects annual savings of more than €100 000, and the estimated savings will increase to almost €1 million per year once all phases are completed. It is therefore no wonder that Pauli Saarenpää and Mika Salmela recommend working with Dynamik to other companies with similar needs and delivery volumes. “They are also easy to work with and very nice people.”