As in virtually every other industry sector, artificial intelligence (AI) techniques are helping cartonboard manufacturers like Metsä Board analyse data and figure out how to optimise production processes to get the best possible result. Tomi Vähä-Ruohola, Technical Development Manager at Metsä Board, talks about the company's journey in AI so far, and what it takes to get from idea to production-ready software.
Our journey began in 2018, when we started to rank the areas where we saw the greatest potential for AI to improve our production processes. Carton manufacturing is already a highly technical process and our factories have very advanced automation systems. With AI, we can go one step further, as it allows us to optimise based on data collected at thousands of points in the production process. Further improving the quality of our boxes means less waste, more efficient use of raw materials and even more satisfied customers.
Accelerating data collection
With sensors collecting data and Artificial Intelligence processing the numbers, we can skip the time-consuming and tedious process of taking carton samples, analysing them in the lab and waiting for the results. The virtual sensor we developed in cooperation with a supporting IT partner predicts the quality value based on the lab analysis results entered at the beginning - the more data the sensor gets, the more accurate its predictions become.
Building a bridge to the future
To do this, we not only had to develop the AI analysis model itself, but also figure out how to transfer the huge amounts of data from production to our partner. Developing this kind of data architecture was something we had never done before, because at the same time we had to make sure we developed a watertight, secure system. Metsä Group's ICT team has been an extremely valuable partner in this process. The end result is a "data bridge" from our factories that we can use for all kinds of analytical applications in the future.
Many applications in the pipeline
A concrete example of a new application idea can be found in the threading process, where ropes and other equipment are used to rethread the roll sheet when the board machine is restarted after a stoppage or a break at the wet end. The threading ropes wear out over time and are located in areas that are not accessible when a board machine is running. Employees at our Äänekoski mill suggested monitoring the condition of the ropes using machine vision and AI analytics to predict the optimal time to replace them and avoid unnecessary downtime.
Know-how and networking for further progress
We use our internal knowledge of the carton manufacturing process to identify applications and our data bridge to obtain data from production and the environment. We trust IT experts to develop the algorithms for us. With this technical infrastructure, we can develop many new ideas for the application of AI and thus continuously improve the quality of our cartons for the benefit of our customers.
www.metsagroup.com
