A modern paper machine produces vast amounts of data, but operators decide how order fulfilment is done. Little is done to use available data to optimise production costs, forecast important laboratory measurements or entire production line efficiency. Important decisions are made ignoring solid support material.
Paper machine automation is extremely good at driving the most important quality parameters to target while producing the batch in minimum time, but ignores business targets. It is unable to consider the entire production line and the overall production economy end-to-end.
We need to balance quality, production costs and runnability while keeping the customer promise. As quality, especially in the board industry, is often customised and increasingly important, automation and business systems need to exchange much more data. Each production run being different, advanced optimisers need much more information of the business situation and used raw materials used to be able to optimise properly. On the other hand, business systems need more data from automation to be able to optimise and control operations throughout the complete production line.
This is where the enormous volume of data comes into play. Based on data, developed measuring technologies and machine learning, AI-based modelling will adjust the optimisation model and make predictions. It continuously learns from the model and actual data. Tieto’s AI projects in the industry show that tangible improvements are possible after only weeks training an AI.
The first concrete step is to provide instruction to machine operators to complete planned customer orders with optimised quality, as cost-efficiently as possible. The next step is to continuously and automatically re-trim production to further optimise order fulfilment and the use of already produced material.
Optimisation always starts with measuring and continues with analysing, modelling and forecasting – in this sequence. Modelling is nothing new in the paper industry but modern technology now provides standards for storing and accessing all data, enabling huge cloud-based computing power and bringing modern AI-based technologies into common use. Ecosystems – like Valmet and Tieto – enable deep integration across system boundaries. New doors are now opened.
An automatically optimised machine chain will run production more economically and avoid both under- and overproduction. In future, operators can work most of the time in the field supporting the production process and intervene only when decision-making is needed.
True, the road to fully automated, self-learning production chains is long and somewhat winding. Measurements are already there, what needs to be done is analysing the results and create a model, teach the system for accurate predictions, and only then can you optimise. There are no shortcuts.
For many companies, the first step is to collect all data under one, commonly accepted interface. Production systems are rather intolerant to wide-ranging data searches, so data should be copied into a data lake that can hold vast amounts of unstructured data in different formats for further use.
Tieto’s industry expertise, integration competencies and process management knowledge help you create a platform on top of which specific solutions can be built to make use of the data in the lake. Tieto has previously designed data lake solutions for hospitals where important new insights for better processes have been achieved.
If this sounds like science fiction, it will not in a few years’ time. AI-based solutions are advancing rapidly in all industries, and Tieto has several references of successful implementations. Wouldn’t it be time for you, too, to start using the power now dormant in your existing data before your competition does?
Get familiar with Tieto's pulp and paper solutions on this website.
Read more about our Data and AI services.