EPISODE 7: How software pulls everything together

Sarah: Hi and welcome back to Listen LiSEC, the podcast about glass processing! My name is Sarah Hummelsberger, and I am a part of the LiSEC Marketing team. And across from me is Harald, who will introduce himself.

Harald: Hi, my name is Harald Miksch, and I am a product manager here at LiSEC.

Sarah: We are recording as always at our production site in Lower Austria. Our studio is situated inside the Training Center, so you might hear an IG line moving in the background. Speaking of IG, you’re currently working on an interesting IG project – you’re renovating a house built in the 70ies. And didn’t you tell me you’re replacing the windows – actually, only the glass panes?

Harald: Yeah, right! The window frames are still okay, but I want to improve the insulation value, so I’m switching to thermoplastic spacers. I only placed the order with my glazier last week.

Sarah: Cool! So far in the podcast our focus has been on machines, but I know your area of expertise is Software. Can you walk me through the process that happens in the software now you’ve placed your order?

Harald: Ah - Sure! Let’s start with the email I sent last week. There, I listed the dimensions of every window, and the specifications. And usually, this is what happens next: The person in the office enters the information into an Order Management Software. In my case, that’s 16 panes with 6 different dimensions.

Sarah: And will you get them in time before your vacation?

Harald: No - Unfortunately, I have to wait a little longer. You see - When I requested my preferred delivery date, their capacity planning checked the machine load for the next weeks and immediately told the clerk that they would not be able to meet the date. Instead, they suggested the earliest possible delivery date – a little time after my vacation.

Sarah: Oh no, that’s quite a wait. Did you request any special glass types they do not have on stock?

Harald: Not so much the glass, but the coating is a bit out of the ordinary. Based on their stock management data, they could directly tell they need to order that type of glass. And, that kind of information is provided by a Purchase management system, which verifies the stock level. So - Once the glass on stock falls below a certain threshold, the system restocks a preconfigured amount from a default supplier. That means my order must have exceeded their current stock level.

Sarah: Did they offer you a partial delivery?

Harald: Ya - They did, but I decided against it. It would be a hassle for the glazier [gleisiea] to have to come twice to replace the glass in the frames, and I don’t have the space to store the glasses, either. So, they adjusted their delivery planning to make a complete delivery – regrettably, after my vacation.

Sarah: What happened once you confirmed the order?

Harald: This is the moment where the reservations in the capacity planning – due to my request - are changed into actual production orders. These orders contain all necessary work steps, which are then scheduled on all involved machines. Usually, that scheduling is done from the delivery date backwards to make sure everything is finished in time, but nothing stands around on the shopfloor consuming racks and space.

Sarah: How does the software support the production?

Harald: A major function of the software is actually optimisation. This happens first during cutting, where orders - selected by the operator - are optimized together on series of stock sheets - for optimal yield and minimal waste. In a more automated production, the optimization algorithm may even decide to include panes from later production orders - just to ensure a better cutting result, even if that means they take up space on a rack or in a sorting buffer.

Harald: A second optimization happens during the IG assembly, where the elements are produced in decreasing size – so they can be stacked on an A-rack. This is then called a delivery rack optimized production.

Sarah: And after all this optimization, how does the glass pane find its way through the production?

Harald: Hah - The foundation for that is identification. In the software, each pane has a barcode identification which can be printed on a label or sometimes even lasered into the glass. Every time a pane arrives at a machine for processing, the operator scans this barcode. Then a visualisation - like a loading display - tells the operator how to put the glass on the machine - and also what exactly needs to be done.

Harald: If there are multiple machines in a line, a line management application does supply the work step information to each machine. By doing so, the operator does not have to scan the barcode at each and every machine.

Sarah: So the software provides optimization and identification – what about keeping track of all the moving pieces?

Harald: Well, there are two basic status messages. For each successfully completed work step, the operator - or in an automated production the machine or line management - sends a ready message to the capacity planning. This way, the software system can keep track of the production progress.

Sarah: And what if – worst case – something brakes?

Harald: Yeah - That’s not as bad as it sounds! Since glass is a material that tends to break frequently, there is a process for that. In the case of a breakage, the operator initiates a so-called remake message. This restarts the whole production process - for whatever has broken - starting from cutting. And - This is a major advantage of the dynamic optimisation algorithm I mentioned earlier: it recalculates the upcoming optimisations to include this remake as soon as possible.

Sarah: That reminds me! Earlier, when we talked about the cutting optimisation, you mentioned a sorting buffer. What is that?

Harald: Hmm - The buffer is the key for a fully automated production. You know - From the cutting, the glass panes arrive in the sequence that ensured the least possible glass waste in the cutting process. The IG line requires the panes in a different sequence, though. Here the goal is to stack the panes on the rack at the end of the line in an optimised way – a completely different sequence! Calculating this hand-over is one of the key functions of the buffer software. This is not limited to IG production – the buffer can also feed a lamination line or a toughening furnace. And - This can get quite complex! So, you see how the buffer and the buffer software automate the logistics in the production.

Sarah: Impressive! If you have a fully automated production, what’s the role of the operators – or even the production manager?

Harald: Well - Even in a completely automated production there are still tasks that require human action. Those are then mainly tasks which are less repetitive and more sophisticated, like maintenance or trouble shooting. The production manager’s role, is to monitor the production from a control centre and to intervene in case of a machine breakdown or something similar.

Sarah: But is that all a production manager has to do? Should they not also look for ways to improve the production, based on all this data they get from the machines?

Harald: That’s actually a good point! State-of-the-art productions generate a lot of data that can be analysed with appropriate software solutions to derive the best measures to increase the efficiency, improve quality, and reduce down times.

Harald: But data is not only generated in the production! Any system, no matter if it’s a business solution, production planning, or manufacturing execution software, generates data that can be collected and aggregated into so called data-warehouse systems. These systems are able to calculate KPIs across processes, which show insights that would not have been apparent from the raw data.

Sarah: Those have been a lot of buzzwords, can you give me an example for a real-world application of such systems?

Harald: Sure! A data-warehouse system answers questions like: Is there a product group that tends to create more remakes? How many orders did we deliver on time in full… in a specific region for instance. So if you have delivery issues, you figure out common denominators to identify the underlying cause, and derive appropriate measures.

Sarah: That sounds quite fascinating. And I am sure, there are many exciting developments happening in this area. What will the future bring?

Harald: Well, the first thing that comes to my mind actually has to do with data. As a production manager gains more and more experience, they know how to deal with certain situations. The same is true for Artificial Intelligence algorithms. The more data you feed them, the better equipped they are to react to non-standard situations. The algorithms do not only control, but constantly inspect and adapt to the current situation. In that case, the human production manager only adjusts the parameters to guide the algorithms decisions.

Sarah: Oh, that’s a major change! But surely that’s not all?

Harald: Of course not. A second innovation is the move from the push-principle to the pull-principle in the production area. Nowadays, a lot of productions try to produce as much as possible in a given time. They increase the output, regardless of their capacity to deliver the finished products. Using the pull-principle, you would do the delivery planning before the production planning and therefore only produce what you can actually deliver. By doing so you decrease the needed space in the warehouse and stay more flexible if order changes occur.

Sarah: And what about the humans? Any improvements for them?

Harald: That’s the third focus: the user experience. You know - Industrial software has the connotation of being hard to use, complicated and training intensive. But - People growing up with easy to use apps on their smartphones expect also industrial software to be as intuitive as the software they use in their everyday lives. So the user experience of industrial software will develop drastically.

Sarah: Quite a cool outlook! It seems software will bring quite a lot of changes to flat glass processing and make life easier.

Harald: Yeah, because basically, a shopfloor is only a set of machines until you add software – then it becomes a production.

Sarah: It’s been quite a journey, going through the aspects of flat glass processing with you. Today’s episode concludes the first season of LiSTEN LiSEC. Thank you so much for this cool adventure and thank you to all our listeners for following along!

Harald: It’s been a pleasure! Thank you very much for having me!

Sarah: This first season has laid the groundwork for what is about to come: detailed talks with experts and deep dives into what makes our industry so special and interesting.

Harald: We’ll keep bringing cutting-edge technology news to our listeners!

Sarah: Hear you in the second season.

Harald: We’ll be back soon. Shopfloor stehen ein Haufen Maschinen – nur durch die Software wird daraus eine Produktion.