Get yourself connected

Earlier this year, Albrecht Baumer launched Sophie, a modular production control and data management system for polyurethane slabstock processing. The system offers much more, though – and could be a jumping off point for companies wanting to raise quality, and improve efficiency as well as understanding their processes.

Some of the most innovative ideas to improve process transparency in polyurethane slabstock processing have not come from the companies who make the foaming machinery. Rather, it is the ancillary equipment makers that have been developing ways to stay ahead of the competition by adding value to the foaming process in different ways.

Sophie is the name for Baumer’s latest offering in this area. The system integrates up to eight modules that monitor and optimise different parts of the polyurethane slabstock manufacturing process.

Like other systems, it is designed to help reduce scrap, while improving efficiency and transparency by gathering data. It also applies artificial intelligence to the large amounts of data it can capture. Sophie interprets this information to spot links between the different aspects of foam making, storage and cutting. It can also recommend actions to production personnel.

Use the AI

‘Intelligent machines open up new possibilities in data collection,’ said CEO Jan Henrik Leisse. ‘Things which in the past were dealt with by gut feeling can be analysed now, so the processes become better understood. If you can measure it, you can change it. That’s why we are developing intelligent control systems as well as machines.’

Project engineering director Michael Jenke said that until Baumer became involved in gluing with its purchase of Lamit, it was not fully aware of the scope for process improvement. ‘When our machines cut the sheets from the blocks we didn’t care about the orientation of the stacks, that was not our problem,’ he explained. ‘But it becomes our problem when the foam is processed further.’

Baumer looked at all the steps from block production to storage in the development of the system. At its heart is Sophie Data, which links all the data capture units, as well as other cloud-based or customer databases. It takes the raw data they provide, analyses it, and feeds suggestions back to plant operators.

 ‘Measurement is a good place to start to explain how Sophie works,’ said R&D director Christoph Moisel. ‘The measure module measures the block as it is produced. It takes more than a thousand data points at hundreds of cross sections all along the block length. This is converted into a 3-D model and gives a profile of the long block. It enables the producer to precisely understand the geometry of the raw block, as well as failures such as shrinkage effect

Check the spec

Moisel added  that if the block is out of specification, the plant operators can be warned that they need to change production parameters. The measure module also gives the weight of the block. 
‘We can measure the weight of the long block, when it is newly made and at different times during storage,’ he said. ‘This enables the fresh density, the density that was planned and the real density of the finished block to be measured.’

Sophie compares this with existing data, and tries to visualise the data against the plan or production history. This allows conclusions to be drawn about the recipe and enables it to be adjusted if necessary. It allows quality to be increased with experience, he added.

The inventory module provides a comprehensive inventory of tank storage, curing racks, crane and short block storage. After the long block is cut, Sophie’s inventory module can give the shorter pieces unique identifiers. These contain all the process information, and follow the blocks on their journey through the block store and out the other side.

Using the identifier, monitoring block temperature in the storage system can help foamers produce more consistent products. ‘As soon as a block is moved from one position to another, the data is copied to the new position,’ explained head of software development Volker Czymek.

‘This can be useful if you are in a cold or hot country. There could be a different curing process for blocks in the summer. In the wintertime, blocks on the outside of the store may cure much faster than those in the middle, because they are not insulated by other blocks.’ This subtle change in curing may lead to different foam properties and Czymek said foam makers may find that useful. 

Cure problems 

Additionally, the dimensions of the blocks can be checked as they leave the store and compared with the original dimensions. Changes in size between raw and cured blocks can be measured, recorded with time and analysed. This makes it possible to check how the foam behaves from pouring, through curing and storage and into the production process or transport. 

After leaving the block store, blocks are monitored by the Produce module. It is designed for small production flows, and can handle processes such as cutting a short block from a long block to make sheets that will go to a CNC machine. 

Identify is another feature being planned. It will follow polyurethane from long block to short block and into the final product via bar code, RFID or similar technology. This could make it possible to give traceability down the chain to the final customer. 

Baumer foresees that even if the final product is complex and consists of several layers, such as a sandwich mattress, data would be available for every single layer. This could be helpful in dealing with customer complaints or recall campaigns.

Prepare properly

At the moment, though, the last stage in the process is Connect. This talks with the customer’s ordering system. ‘We can build an interface between it, different machines and other customer databases,’

Czymek said. It connects with Sophie’s Prepare module. This module enables block production planning without the orders being overwritten.

While these are some of the ways Sophie can help foamers day to day, Moisel believes there are other, less obvious, ways that AI can help foamers. ‘We can collect all this information, link it together and try to get useful analysis out of this in the Optimise module,’ he said. 
Data collected in real-time from the plant can be combined with the experience of users and machinery designers.

‘This can give the operators guidance about how to react to a foam quality problem,’ he added. ‘Or, perhaps, if we detect temperature differences in the curing rack, Sophie could advise on the best way to align the blocks. It could suggest which blocks to move to achieve uniform curing across the process.’