Improving machine vision performance at tissue mills through better use of data

There is today a wide range of automation equipment and measuring tools available for tissue machines. All of them provide important information about the process condition but are often operated in independent clusters. As a result, it can be challenging for the machine operators and production planners to get a holistic view of the machine’s performance.

Furthermore, data analysis becomes complicated as it requires special skills to understand correlations between pieces of information available in separate systems, leading to delays in problem-solving and decision-making.

Valmet’s solution: faster handling of relevant data through one integrated system

For decades, Valmet has collaborated with its customers around the world, offering state-of-the-art papermaking technology, automation solutions and related services. Over the years, Valmet’s experts have acquired thorough know-how on papermaking and its numerous processes.

To solve the issue of siloed data and improve efficiency in handling it, Valmet has integrated all data from its equipment into the same system.

Gathering process tags, web defects, profiles, web break logs, and trends in one location makes it easier and faster to find correlations between different process parameters, which is especially useful in different problem-solving situations.

For a concrete use case, let’s look at edge cracks, a phenomenon known to many. When all relevant data is visible in one screen, the operator can drag for example machine speed from the user interface and check possible correlations with edge crack frequency. It can easily be seen if the appearance of holes is by any means correlating with the chemical dosing level change.

Towards autonomous operations

Even when data is collected and analyzed in one place, problem-solving still requires special skills and manual work. The next logical step would be to automate troubleshooting and decision-making, taking tissue-making towards even more optimized and autonomous operations.  

In an autonomous mill, the system can monitor its own performance, leading to increased efficiency in conclusion-making. Imagine finding the root cause for the already mentioned edge cracks in some hours instead of suffering from web breaks or poor quality for months. There’s a huge savings potential.

Educating Artificial Intelligence

Artificial Intelligence, AI, is seen as one of the most potential solutions to increase autonomy in tissue mills. But it has its deficiencies.

The paper manufacturing process is complicated with many possible unexpected disturbances – meaning that also AI can become less intelligent if not educated with enough relevant information.

To benefit from AI on a big scale, we’ll need to be able to see and observe the production process through a huge amount of process data and information. For most mills that require new measurements, a higher level of meaningful process tags, advanced camera technology, and intelligent applications helping the operators make sense of the data to resolve issues faster and independently. “As an example of such an application, I could mention the Valmet Centerline Advisor which is an advanced application for comparing the current process status to the centerline of the best historical performance”, says Tommi Leander. Unfortunately, there are no shortcuts on the way to becoming more autonomous.

In sum, the more data we have and the better it is integrated, the more intelligent the production line will become. The more intelligent the production line, the bigger the benefits related to safety, efficiency, profitability, and quality will be.