0:08 Welcome and hello also from my side, I’m Daniel, co-founder and co-CEO of Synera, and yeah, I’m super thrilled to be here today and to talk about, we call it process automation. I would say it’s maybe a different term than computational design, but we also tried a lot of other terms, so we started with generative engineering, then connected engineering, now process automation, it has all the same purpose. Our goal is to accelerate product development, especially for mechanical engineering and for mechanical products. But before I jump into our product and our offering and the case studies, I would love to share a little bit, yeah, the industry view, what we see at our customers and partners.

0:51 So our engineers, when we look at automotive, industry space, industry aviation, they all work pretty sequential, we also saw a lot of these sharts already this morning, so we have the design department, we have the simulation department, and maybe we have optimization, costing, sustainability, they’re all like queued up in line. Then depending on the size of the company, we have a lot of different departments, maybe even different locations, so it’s very hard to connect them. And we already have seen it, there’s a lot of different expert tools available in the market for all the expert disciplines, so you need experts to use these software and also to combine them to have a great product at the end. And last but not least, you also need the experience, you need the engineer who know how to deal with all the things and combine this, so it’s a pretty tough challenge.

1:46 And the question is really, what can we do if our goal is to accelerate product development, what kind of strategies do we see in the market? So the first approach you see, it’s pretty obvious, it’s like, it’s a manual task, let’s increase manual work, let’s hire more people, train them more, and yeah, make, get the [ __ ] done. Or the second one we see, it especially in the automotive industry, is like, okay, let’s deal with engineering service provider, let’s outsource the TDS work that nobody want to do, and best case we use best cost countries to to lower the price, use our bargaining power to make sure that we don’t have to deal with these tasks.

2:32 The second approach we see is, and this goes more in the direction of computational design, is automation, and there’s a good thing that all the expert tools, they have automation already in their DNA, but you need more or less another export to automate these tools. I don’t know, for example, we have Ala with with TCL, we have ANSYS with RPDL, we have a Ctio with Weba, luckily we see more and more software V pushing to Python, so maybe some some light at the horizont, but yeah, let’s hope for that. And last but not least, if you really want to connect the silos and do it like a full product development processes, the tools have to speak together, and this is even the next challenge.

3:22 So we thought about what could be a perfect world where we as an engineer really enjoy our work and free ourself from tedious work, like repetitive work that nobody really is enjoying. So we thought, how cool would it be if we are able, every engineer on its own, to digitalize their working steps, their sequence of work, that they can ever reuse, and the best case already share with their co-workers and scale inside of their organization. So with this in mind, we came up with a new programming language especially designed for the mechanical engineer, so we developed a low code language which is like a visual representation of a programming language.

4:08 We have already seen some Grasshopper slides already this morning, so it’s very similar to to Grasser, but with the focus of mechanical engineering, not so much of the architectural or product design. And you will see in a second how this is working, how you can digitalize your work steps into to a workflow and how you can reuse it and scale and share it with your co-workers. And this is also the biggest difference, and I think it’s probably for all the computational design products, is that we are not focusing on the final product, we are focusing now on the steps to the final product, the workflow itself.

4:47 Because if you are concentrating on the workflow and not on the final increment, for sure the final increment is the product we all need at the end, we want to commercialize, but if you’re focusing in the development on the workflow steps, then you can reuse it, you can scale it, you can share it with cloud compute, you can create better products in in days and not in months or years. So let’s first understand a little bit how low code works and how it looks like. So this is a canvas, a very simple workflow, but sometimes simple workflows are even making more money as we think, so it’s like connecting the nodes with wires and data flows from left to right, and every engineer is able to digitalize their processes, and here it’s pretty easy, open a lot of geometries, extract the volume, and save it to an Excel file.

5:43 But engineers are living in a more complex world, right, we have more things to deal with, and our goal Atara is not to reinvent the real, so we do not want to create a new solver, or a new geometry kernel, or a new, I don’t know, CFD instance, we believe in that there are already so great tools available, and we have seen already so many of it. So you see the the logos on the left we are partnering with, so the idea is really to grab these kind of technologies and unify them under this universal low code language, so that the engineers can use the tools that they love to use and they they really want to use in the background for automation across the different silos and across the different, yeah, discipl.

6:25 So we already have talks seen from CDs, we will see later Rafinex and Simscale, Navasto tomorrow, so a lot of very cool integrations are already available in Synera. So how does it look like now from a connected engineering approach? So this is a very short demo, but I think it shows how you can, for example, combine computer AED design, like CAD, in this case it’s a node that is able to remote control Seens and X, and then you have, yeah, exposed some parameters in your parametric cut model, and you can change them. Then the data flows to preprocessing nodes, and at the end you have a solver node, in this case the customer used OptiStruct, and again then all the results get, yeah, packed together, and interesting is, I would say eight from 10 customers ask us to have Exel at the end, so it seems that Exel is still the winner.

7:26 But yeah, let’s see, so the so this is the technical presentation, first of all, of low code and what’s all about, and I know that you guys want to see real world products and real world projects, so I tried to, yeah, find someone that I’m able to speak about it, and luckily there was one customer presenting a research paper on a feasibility study last year during a design symposium, so I’m super thrilled to give you some insights from this study. So the study is like feasibility study on time saving potential of automated workflows in the early design stage of bus body structures, so the bus already gives you a glimpse who could be the customer.

8:04 So this was a feasibility study of the T and Min truck and bus, and they wanted to benchmark manual work against computational slpress automation in product development on a larger scale. So what was the project all about? So they they producing buses, and their customers have a lot of different requirements, like, I don’t know, I want to have this feature, I want to have this length, this access, and all the different requirements that the customers normally want. And what Min is building is that they have a huge library of these kind of segments for all the different purposes, and you can think about in the phase one that they more or less puddle these together.

8:54 So from a computation design perspective they created these kind of segments once and then they could reassemble and reuse them in all their sequences, and then they even wanted to push the limit a little bit further. So they thought about, okay, what happened if we have a combination where we have to reinvent, where we have to make adjustments which the structure is not capable of? So this is the phase to, where they also introduced topology optimization, or somebody called it generative design, these kind of algorithms like Sim Bzo and all the stuff, you know, to generate lightweight structure based on requirements.

9:35 So they were able to have a holistic process where they could say, okay, I want to have this, yeah, assembly of segments, I want to redesign this segment, and then the whole Stu starts from from the beginning, and then you can learn it and put into the database again. So here’s the the same, but here’s more like the the text run, so we have typical, what we need is on the on this column, we have the automated steps and we have the manual steps, and the automation step they did is written in bulk. So that means you have typical pre-processing work, you have to call the solver, you make a post processing, you have to interpret the results, you have to get some design changes from this result interpretation, you have to prepare the simulation optimization, and you have a feedback loop to realize all this.

10:25 And the interesting thing is here, that we really want to look on the comparison between the manual work and the automation, so it’s really cool to see, they made a comparison and they really had the stopwatch running, and say, okay, now let’s let’s measure the time for this specific use case, how much time it is to do it manually versus in an automated way when we have the automation in place. And I think the numbers are really really awesome to see, so the first chart is a lead time, like how many time you really need to run the workflow, there are some, yeah, let’s call it manual steps between some sequences, so there is some time needed, and more stunning is the required man hours, so they were able to reduce the manual work by 96%.

11:18 So by focusing not on the segment as a result but on the workflow creating the segments, and this is the biggest difference, and I think this is a huge enabler in all the computational design projects we have seen here on stage. So when you’re focusing on the workflow and not on the part, then really this kind of magic can happen. For sure you have to invest something, and we also have seen it in other presentations, you have to build this kind of automation, right, it’s a lot of work covering all the edge cases, make sure it’s robust, it’s working, and you do not have, like, errors or bugs, it becomes more more like a software product than a than a hardware product.

11:52 So we see they reach, well, they they they got to break even by the run of 77 runs, so it means before 77 runs it was much, well, not much, but it was cheaper to do it the old-fashioned manual way, but after this iteration they got a boost in productivity and boost and efficiency, and for the men ours even a little bit lower, but yeah, it’s 74. I think this is very important to see, because we we heard a lot in the early presentation about return on invest, right, we, it’s, see so many great products and great things, but at the end you have to make sure it makes sense also from an economic perspective.

12:36 And I think especially automation makes sense if you have repetitive work, and make sure you really have repetitive work and not just once a year or twice a year, and make sure there’s a certain complexity in it that you really can eliminate here. If you want to have a further reading here, this is, U, yeah, a paper released last year, happy to share the link, or it’s also written as a course, so it’s really interesting to see how they did this deep, yeah, comparison between manual work and computational design. But again, this was only one project out of many projects, and I think the biggest challenge we have at Synera is that, yeah, we develop the low code language, it’s like a programming language, and everybody grabs it and makes something out of it, and sometimes I go to a customer and I see applications I’ve never thought they will do this with our software, so it’s very hard to to be on spot in certain areas.

13:28 We see so many different, yeah, applications, and I thought, okay, let’s let’s make a selection out of it which we can share, and to give you some insights how you can use computational slpress automation. First of all, what is our customer base? So we are normally, U, working, I would say, 60% in the automotive domain, but also in the space aviation, consumer goods, engineering service provider, all, I would say, larger companies who have really this kind of repetitiveness in their products, and they really want to speed up things and make sure they are getting more efficient. And as you see, this 76% on average is really a huge number, but this is only possible because you’re changing the way how you work, you’re not focusing on the part, you’re focusing on the process.

14:18 Sure, you have to invest in the process, but then you can reuse it, you can share it, and you can scale it. So let’s talk a little bit about return on investment, so we have workflows seen by customers that they have a huge return on investment even if they just get run once, it’s insane, they just run it once a year, and return investment is there, but we have also seen applications where they have to run 10,000 times a day to get a return on investment. So it’s really important to to see, okay, what kind of effort do you have to put in, an effort could be also workflow complexity and how many runs you really have.

14:53 Because it’s also, okay, I mean, I showed you the very first demo with extracting the volume of a CAD model and put it in Exel, it’s very easy and less effort, you can build this in two minutes in Scenera, but if a company uses 10,000 time a day, then it’s already value, right, and this is important to understand. So when does it make sense to apply this kind of technology, and yeah, if you have this in mind, then you will also find your your use cases, your return on invest, and then it also makes fun to see when they get the sparkling eyes when the first time the automation really runs through, and especially reporting is something that everybody hates, so this is definitely a bestseller.

So I selected three different areas, I know that most of you focusing on additive, so I will also show you two additive cases, but I would also love to talk a bit about simulation, as we will also see Sim Scale in a second, and even a hidden champion, I would call it, utility workflows, it’s insane what C customers doing in this domain, but yeah, it’s like building the helper tools that everybody wants to have to to make sure they are not repeating themselves every day. So first one is in the domain of automated FAA, we have seen this in the MN case study already, here it’s a company called A International, it’s a larger engineering service provider, and they’re doing a lot of design verification.

16:22 So they change something in the design and they want to get a stress report very fast, and normally what you see in the industry, especially in this industry, but also in the automotive, is that you have much more designer than simulation guys in the team. So when the designers change something and they want to get feedback from the simulation department, they normally have to wait because the backlog is too long, or they have to make pressure to the engineering service provider. So this was the reason why RLE developed a workflow where they put all the different loadcase definitions, automated them for the chassis development, and whenever they change something in Ker, the automation get pulled and it run, and they got their stress report after a couple of hours, when the whole simulation deck was, I don’t know, 25 load cases has been computed.

17:08 Another customer for paper, we heard from from Brad about the large turbines and and Seen Energy, this company doing paper machines, and I also thought, okay, paper is maybe a little bit old fion, but it is not, because we see paper all over, especially when you’re ordering at Amazon, because paper is also, yeah, the box where you get your your stuff, and it’s insane how huge these kind of machines are, they are like 800 M long producing paper. And they also had the problem that they have more designers and less simulation guys, and how to to have this like rebalance between the the geometry guys and the simulation guys, they also created a workflow which even can the simulation, the designer can pull, and I think this was something, was Trinkle showed us this morning, right?

18:00 We have to make sure that these kind of work sequences, these kind of workflows needs to be very easy accessible, otherwise it’s very hard to scale inside of the companies. But yeah, with this approach F is able to give the designer the power of computing their modal analyst with the report they normally have in in an asynchronous way, without waiting and having a meeting with the simulation department. Another customer for us was a bigger tier one automotive customer, they dealing with workflows in the field of additive manufacturing, so they have multiple application, but at least we are able to speak about that one.

18:36 So they automated their whole, let’s call it, defund process, so getting the design space, making the pre-processing checks, removing all the part that they are not needing, making the topology optimization, reconstructing the topology optimization, make sure they have a nestable design, nest everything, make the support structure, and put this all together in the loop to make sure that they are cost efficient, because at the end, I mean, cost is always what matters, especially in additive, and when you want to bring additive into automotive it’s the number one goal to to be cheaper and focus on the production costs.

Then the last project, because I’ve seen that BMW will also present later, I were able to put this on slide, unfortunately it’s not the automotive project, here I’m not allowed to to talk about, but this was another very interesting, yeah, project together with BMW and their additive department. This is like the, yeah, also sports equipment, supporting the the bob team to having customize spikes under their shoes to make sure they have more grip, more acceleration, and yeah, more success in the competition afterwards.

19:49 Okay, and last but not least, I want to talk about utilities, and for that I would love to show you an additional product from Synerawhich is we call it Synera Run. So Synera Run is the idea of scaling your knowledge to someone else in the company without learning scenario, it’s again this kind of thing, okay, what do we have to achieve to make automation accessible, maybe even for some department that have never touched CAD, for example, we saw people who use it now from the purchasing department, they never open Ktia or NX or making operations, but they would benefit from information coming from CAD.

20:32 So with Synera Run you can build your own workflow with low code, hopefully as easy as possible, and then you can just say, okay, I would love to create everything together, I compile it and make a web application in seconds. So let’s see in this example, I think it’s also a simulation example, yeah, very easy one, some forces, and the guy here, here register one of the input in the low code as an input, and another output here, which is like the stress results as an output. Then, say, okay, publish everything to our Scar Run server, you say, okay, what kind of permission you would love to set up, which co-workers should have access, and then they get a web UI created on the fly, and you see you have an input here, so you can grab any kind of geometry, and then the computation happens on a server hosted on your infrastructure.

21:30 So it’s on your security, and you can now utilize workflows via the web browser even without having all the software installed, making sure the connections are working and all the stuff. So you can really deploy any kind of automation within seconds and make your co-workers happy that they can, yeah, use your work without bothering you, and you hopefully have a little bit more time for coffee in these days to to enjoy and see how people are using your work. So one customer is already using it heavily, this is Purm, they are also in automotive tier one, and they created a utility workflow for part comparison, so they have a special way how they normally do it manually, how they compare different kind of parts.

22:14 And you see already the return on investment, so it’s crazy when you can make it so EAS and so accessible through others, because then your amount of end users is like much larger. Normally we see now factor 10x, so normally we see one person creating these kind of workflows and 10 people really using these kind of workflows, and then you really, yeah, have fun when you try to to come up with a return on investment calculation to make sure that they can also afford the license cost, and it’s a win-win situation for everyone. So yeah, this was my presentation, I hope it was, yeah, interesting for you and you have seen something new, and you are also interested now into loow code and want to have some hands on, happy to have a talk later, yeah, I’m available and also happy to connect on LinkedIn, thank you so much.