#28: What Is the Final Piece to Future-Ready Operations?

Mike Bradford: Currently, what is driving manufacturers towards resilient, sustainable manufacturing are several key industry challenges. First, product complexity and diversity, especially in industrial equipment. We're talking mass customization. Everyone wants their product to uniquely meet their needs and be uniquely configured or designed to solve their problems and achieve their purposes. So, complexity and diversity are growing.

Second, the workforce challenge. According to the National Association for Manufacturing, there will be 2.1 million unfilled jobs in manufacturing by 2030. It’s due to retirement, higher turnover and so forth. So, that's one workforce challenge. Also, higher employee turnover as you don't see employees staying for 20 or 30 years anymore. That combination of unfilled roles and higher turnover has driven the cost of employees up. That also then leads to higher new employee onboarding — longer time, higher cost and just the image of manufacturing — all of those work together to really provide some big challenges in terms of workforce in the industrial equipment industry.

Another one is technical transformation. I mean, new technology is growing. People are talking about digital technology, adoption of things like digital or virtual twins, artificial intelligence and machine learning, augmented reality, virtual reality, as well as something called IT/OT convergence.

And just quickly, if that's not a term that's familiar to you, Gartner says that it’s one of the leading challenges this year in manufacturing. IT is information technology. Think about things like ERP, PLM, manufacturing operations management, or MES — information technologies that are typically driven out of the front office. OT is operational technology. Think IIoT devices and machines and equipment on the shop floor. The drive to pull those two big areas of information technology or of data technology together is really critical to manufacturers as we move forward. So, that IT/OT convergence is another big part of that technical transformation.

Another challenge is supply chain instability. We all remember COVID and how that affected supply. We were so short on a lot of parts as they were stuck on ships at docks and we had this big shortage. And then we had that bullwhip effect, where we now have too much of some things. When we kind of recovered from that, we started to have conflicts in Europe which affect supply chains. Now, we've got new conflicts with tariffs and potential tariff wars. All of those things have caused supply chain instability. So, supply chains continue to be a challenge.

The final challenge is actually one of the topics of today, and that's sustainability requirements. I mean, customers are demanding that we become more sustainable, and there are a lot of new regulations and rules and laws around sustainability.

Those five challenges are really driving a lot of what's going on in the industry. In order to stay competitive, it's critical for manufacturers to address these challenges by adapting their processes and technologies. This is not an easy task, but if it's done effectively, it will enable you as a manufacturer to shift towards a resilient, sustainable manufacturing operation that will stay competitive in this ever-changing environment.

Mike Bradford: An integrated digital platform is ideal for enabling this transformation — bringing together the entire process from ideation to execution and back — on a single platform. It really provides a lot of capabilities and benefits.

Number one, it drives collaboration. I mean, there's really inherent collaboration because all the data is on the same platform and shared. It reduces time and cost for new product introductions and product changes because of that continuous feedback and inbuilt collaboration. Number three, it provides faster and simpler management of complex products — that tighter tie-in — that tight integration between product engineering, manufacturing engineering and execution. It also decreases duplication of effort. 

Instead, we do what we call model-based systems engineering. So, instead of taking the design model and passing it to manufacturing engineers so they can work on it separately, and then passing it to the execution system so that they can actually execute against it — instead of having three models where you lose sight of changes in what's occurring across those models — we do model-based systems engineering.

As model-based systems engineering is a single model that is shared by all, it really reduces a lot of duplication of effort, and it also increases the speed of problem solving and issue resolution. Again, because now we're working on the same model, it's immediately apparent that if design engineering makes a change, manufacturing engineering can adjust immediately. It's much easier for execution — for the shop floor — to feed back through the engineering cycle when they find issues or problems. It also reduces confusion between engineering and manufacturing. Again, we're all working on the same data, and we're using some of the same terminology, so it reduces confusion. Ultimately, it provides significantly higher customer satisfaction because you have faster time to market, faster changes, fewer quality issues — because of the faster and improved feedback loops.

I'll give you one example. XCMG Machinery is a manufacturer of construction equipment in China. They went to a single platform model, and what it helped them do is improve efficiency in new product introductions and improve efficiency on the execution side, which significantly reduced their construction lead times.

Another heavy mobile equipment manufacturer did assembly verification systematically, virtually, using our solutions. Sharing the same data model with design engineering, manufacturing engineering and simulation and virtualization really helped them reduce design changes by 60%, as they were able to virtualize and simulate first.

So, the initial design was better — reducing design changes by 60% — and it also cut their product development lead time by 50%. It cut their product development lead time and the new product introduction lead time in half. Having an integrated digital platform provides a significant step in digital transformation, enabling resilient and sustainable manufacturing operations.

Mike Bradford: The single biggest hurdle that manufacturers that I've dealt with face in integrating digital solutions across operations is data silos. They've got different apps with different data attributes and different data formats, and making them talk and integrating them is difficult, time-consuming and costly. Even when that integration is in place initially, it's never really done. If you upgrade one product that changes one data silo, now you've affected the integration across any other data silos. So, that siloed environment is the single biggest hurdle that I see with manufacturers in this environment. Because we're a single, integrated platform-based solution, the issue of data silos doesn't exist — single platform, single product suite. 

Give you an example. One of our global power and fluidic equipment companies is implementing this now. They started out with our operations management solution. They're now moving to our manufacturing engineering digital solution. This has helped them eliminate a number of point solutions. That's saving them in software costs and integration costs, and it's enabled them to unify processes and integrate design and execution, which then reduces new product introduction time and improves product quality because again, that tighter link between product design, process design and execution really helps to bring a better-quality product to market.

Mike Bradford: Manufacturing Operations Management or MOM, Manufacturing Execution Systems, and supply chain planning systems, working together, really provide strong solutions in this environment. Let me start out by talking individually about a couple of those. 

I'm going to tie MOM or Manufacturing Operations Management and Manufacturing Execution System or MES together, because a lot of people use those terms interchangeably, but they really are different. MES is actually a subset of MOM. So, MOM includes all of your manufacturing execution system capabilities as well as broader capabilities across things like quality, maintenance, material logistics, line-side material logistics, and so forth, time and labor sequencing, if you're in that type of environment. So, MOM, really ties all those together in a single solution. The true advantages of the DELMIA Apriso Manufacturing Operations Management solution are these several key benefits.

Number one is the breadth of the product. It provides simplicity for your IT people — less products to manage — and also for your users. Because if you look at someone in the quality department, well, they need to access the MES information to see what was made and how it was made, and what tests were done, and so forth. They may need to access the materials management system in order to see where the materials came from, where they were stored, and how they were handled. They may want to or need to access the maintenance system to see if there was a maintenance problem with a particular machine. Is that why we're having some quality issues from that machine? By tying all that together in a single product suite, we've simplified that quality person's life as well, because he or she now goes to one system to check all of that, instead of having to jump from system to system. So, that breadth is the first big advantage that we provide.

The second is configurability. We provide a low-code configurability environment. We are a business process management-based low-code environment. We have a library of over 1,500 business components that you put together using our low-code process management solution to create a solution that fits your needs. Because, let's face it, manufacturing is different from place to place, from company to company. We all, as manufacturers, have things that we do that we feel are unique, that give us a competitive advantage, and you want to be able to enforce that on your shop floors and also continuously improve those processes.

Configurability really supports that. And that's really evident if you look at our customer base. We have a number of aircraft manufacturers — large, slow build, very complex products — we're actually one of the leaders in that market. We also have a number of consumer goods customers, for example, in beauty products, companies like L'Oreal and others. But now you're talking about a process industry — much higher speed, much higher volumes, a very different manufacturing environment — and yet we're able to support both because of that configurability.

Third is global capabilities. We have unique capabilities in global process management that no one else that we're aware of has. I've not seen anyone else who can do this. We can take that configuration that we just talked about and share it across multiple plants. You can give the plant some level of autonomy to adjust it for their unique needs, but also you've always got control over that, governance over that, and visibility into that, so that you can still roll out new and improved business processes across your entire solution, across your entire enterprise. So, that ability to not only hand off a configuration, but actually govern that configuration is very unique and is very powerful in a multiplant environment.

A fourth big advantage is that we are integrated with design, so it's a single source of truth. And that goes back to what we talked about with new product introduction advantages and so forth.

Finally, we have added some really good modern technology. Things like augmented reality, to help with assembly instructions, quality instructions, maintenance guidance, so those capabilities from an operations management solution are very, very powerful and make us unique in the industry.

On the other side, the supply chain planning piece, we have two products: DELMIA Ortems and DELMIA Quintiq. With those products, we provide full supply chain planning capabilities across your entire time range — going from long-term things like demand planning and supply planning, where you're looking very long term, into mid-range planning, with things like master production scheduling and workforce planning, to hour-to-hour, minute-to-minute, plant and production scheduling. Also, in logistics planning, if you do your own shipping and logistics, we've got capabilities to help you minimize routes, CO2 consumption, time to get your products to market, and so forth. So, full supply chain capability is one advantage on the supply chain planning side.

Second is the multiple user-defined constraints. You're not tied to — when you're doing your scheduling — you can only look at this. You can define multiple constraints, prioritize what's important to you when you're doing your planning and scheduling, and the system will honor those and will schedule accordingly.

And finally, again, the full integration. Back to MES, MOM for fast correction. I always say you can create the perfect plan, and 10 minutes after it hits the shop floor, it’s no longer perfect because something has changed. You've had to scrap material, an employee hasn't shown up, or a machine is broken down. Any one of a number of things can happen. But if you don't have that fast feedback between your planning solution and your execution system, you can't address those changes quick enough to remain competitive. So, that link not only the individual capabilities of our MOM solution and our supply chain planning solution, but the integration between those two can really help manufacturers tackle some of the real-world challenges that they face.
 

Mike Bradford: I want to give a couple of real-world examples of where we've helped companies with their operations. We had a new customer with a very antiquated system at one of their plants. They were manufacturing engines. Every engine down their line was unique. They did between four and 500 engines a day, and no two were the same, so it was very complex. We talked earlier about complexity and diversity — they’ve very diverse, complex products. So, they had an antiquated system, a complex process and complex products. They really wanted to standardize, not just within that plant, but across their entire enterprise.

They used our solution to simplify and increase visibility in their process. What happens now is they get their sequence of operations from their ERP system into our solution. We, at the first step, begin to enforce that sequence and enforce all the associated rules. So, we tell a material handler what kind of engine block to put on a pallet, and it's an RFID-enabled pallet. So, they scan the block, scan the pallet — those are now tied together —  and it validates that it's the right kind of block for the finished engine that's going to produce with no quality holes and that sort of thing. From that point forward, every time that pallet enters a workstation, we scan that RFID code on the pallet, it ties us to that type of engine, and we give the operator specific work instructions for that configuration of that engine. It simplifies things for the operators — they don't have to dig through a bunch of books. We do things like pick a light line side for fasteners, so they don't have to think about which fasteners to use. We feed that information to feeder lines so they get the right components in the right order for each engine as it goes down the line. Extremely complex, but it's all tied together through our Manufacturing Operations Management system. 

The result is, that through that better integration of visibility as well as automation and error proofing, they've increased their throughput by over 25%, while at the same time reducing their internal quality claims by 90%. It's a huge number, right? They were shipping good quality products, but they were doing it by brute force before, so many people were involved to make sure no bad quality got outside the plant. Reducing those internal quality claims has freed up those people to do much more valuable work for them. It's part of the throughput improvement that they've seen.

It's interesting because typically, throughput and quality can be conflicting goals. In other words, how do you improve throughput on a production line? Well, you take things, you take steps out, you remove tasks. How do you improve quality on a production line? Well, you add quality checks, you add things like that in the line. Those two can be conflicting goals. We were able to enable this customer and help this customer to increase throughput and improve quality at the same time. That configured variation of that solution is now in place at over 30 plants for this customer, and they're rolling eight more plants out this year. They continue to grow and use the product on the supply chain side. I'm giving two examples so they're not tied together, necessarily. But if you think about it, the more these are tied together, the improvements grow exponentially. 

We had a customer called COMEZ. They make specialized machinery. They make machines for narrow fabric production. Think about things like face masks and that sort of thing. Anything that's cloth — they don't make the face masks — they make the machines that make the face masks. When I think about what happened a few years ago, as they began to implement our DELMIA Ortems planning solution, they were planning with Excel and paper, which means you got a lot of data translation. You’ve got to take data from one place, put it into Excel — there are inherent errors in that. I read a statistic many years ago that said the best accuracy you can expect when you're transposing data from system to system is about 85%. So, immediately their schedule is going to be 15% wrong, because the data going in is going to be 15% wrong. They were manually doing things — manually transferring data, running on Excel — it was taking them a tremendous amount of time to build the schedule, and when things changed, they couldn't adjust quickly. We helped them solve two problems simultaneously. Because they were just rolling that out when COVID hit — think about what companies did when COVID hit — all of a sudden, everybody wanted to make masks. So, for these narrow fabric production machines that COMEZ made, the demand almost doubled overnight. There's no way they could have managed the scheduling of that increased volume through their plants. They could not have done it if they'd still been on Excel.

DELMIA Apriso enabled them to meet that increased demand. Now, once that bottleneck was through and they got back to more normal production, it also helped them in all their other scheduling areas. They not only were able to meet that double demand, but they also, once things got back to more normal, saw a 30% reduction in forecast errors, an 8% increase in production —because they're just planning and scheduling better — a 27% reduction in production delays, again, because now they're not waiting materials or waiting for a sub assembly or, you know, all the things that can happen with a poor planning environment. And they saw a 20% reduction in setup times. Now, remember, you can define multiple constraints when you're setting up your planning solution with DELMIA Ortems. So, setup time was one of their constraints, so that it set the schedule — the build schedule, the subassembly schedule, and that sort of thing — to minimize setup times, to minimize changeovers, while at the same time meeting their demand. They were able to reduce setup times by 20%. All those things together — these are just two examples of many — where these solutions individually provide tremendous benefit. Then, when you put them together, the benefits increase exponentially.

Mike Bradford: Now we need to talk about data and its central role in manufacturing. Manufacturers should be able to leverage data more effectively in improving decision making, enhancing overall operations efficiency and ensuring quality execution.

In addition, data is critical to that concept we spoke about earlier: IT/OT convergence. A critical component of IT/OT convergence is what's called a unified namespace, or UNS. This acts as the backbone of integration, consolidating data from ERP, MES, shop floor devices, and edge devices into one real-world data ecosystem. That ensures all departments and systems work with the same data and use a common language, eliminating silos and enabling faster, more confident decision-making across your operations. Data is critical, and by tying it all together, it really enables speed, accuracy, improves quality and so forth.

I'll give you one customer example of the value of leveraging data. It's a company called BorgWarner United Transmission Systems. Now they're a T&M company and they make automotive transmissions, located in China. They're really a digitalization pioneer for BorgWarner, and they selected DELMIA Apriso as their manufacturing platform in order to establish standardized operations for global manufacturing sites and to share important critical data — both within the sites and across sites. They say digitalization was a goal of theirs, and they did have partial digitalization before, and this is their words, ‘but it was in silos’. Remember what we talked about? The problem of having data and systems in silos? So, they had data that was in silos.

DELMIA Apriso broke through all the data from these different processes and silos and pulled it all together. Before, when all this data was in silos, engineers often spent four to five hours to integrate the different data silos and come up with valuable reports and analytics. With DELMIA Apriso, they can now get those same results in minutes — a 90-plus percent reduction in time for what's really a quite expensive resource: Your engineering resource and the data.

Not only a reduction in that cost and time, but improved speed to solutions and speed to decisions, because now they've got the data more quickly and it's available to them more immediately. So, they've not only digitalized and improved their manufacturing processes, they've also united the data sources to improve visibility, decision-making and to more rapidly adjust schedules based on supply chain issues.

It's also helped them to improve quality because, again, that immediate data visibility and feedback enables them to quickly respond, to quickly identify and respond to quality issues. This has helped them to achieve data transparency, so that they can share and transmit data in a timely manner across — within a single plant or across multiple plants — and they are now sharing this solution across three plants, and will continue to increase that. Based on big data collection and analysis, DELMIA Apriso has really helped them focus more on lean improvements, which again improves processes, quality and so forth. That's where data can't stand alone. Data in silos is a problem, and this is one way that we help customers to solve that problem.

Mike Bradford: In order to help our manufacturers optimize operations using digital tools, I've got some simple recommendations. I’ve got a four or five-step process that I think can help with that.

Number one, evaluate your costs. We do something called a value engagement, where our consultants can come in and help you with that. Whether you do it with us or alone, you need to go through your processes and look at where your greatest costs and opportunities are. Where's your low-hanging fruit? Where are your biggest problems? Where can we see the biggest bang for our buck when we make improvements? Because you want to know the value of what you're doing. Typically, manufacturers are very cost-driven, so you need to understand and know where our biggest problems — what they are costing us and what’s the value of solving those problems. So, number one is to evaluate the costs.

Number two is to evaluate the tools you have. Look for old systems that don't really meet modern standards and don't provide modern capabilities — can't do things like simulation, digitalization and virtual twins and so forth. Look for standalone solutions that aren't integrated or are very difficult to integrate with other systems. And look across, even though all of those solutions have the capabilities you need for digitalization. Things like digital twins, virtual twins, simulation, XR capabilities (XR is augmented reality and virtual reality), big data and artificial intelligence — do they have these capabilities? Because those are all important in this process.

Number three, start small. Pick a problem you've identified that you know is costing you. Define what your to-be-state is — how you want this to look after you've solved that problem — then fill the gaps and tool capabilities. Because remember, we've already evaluated our tools, now we know where the gaps are. We know where we're short. So, fill those gaps, and that may mean acquiring new tools, new software tools, new capabilities. 

Define and implement your solution. It may take time, so do it in small steps. Once you've done that, move to the next problem, because remember, success breeds success. When you solve one problem, other people in the company are more anxious to have you address problems that they may be identifying. You're learning more and more how to solve those problems, and you're developing and acquiring better and better tools to solve those problems.

In terms of optimizing current operations, I think that fairly simple four-step process can be really helpful and valuable.
 

Mike Bradford: As we increase the understanding and use of artificial intelligence and machine learning, the transformation and impact on manufacturing operations will be truly transformative. Multiple areas will be impacted — just a few examples:

Number one: The one you hear a lot now is predictive maintenance. Being able to not only detect it's time to maintain this machine, and these are the steps we need to go through, but by using IIoT devices and tying them to artificial intelligence and machine learning, you start to notice vibrations, noises, changes in speed — all the micro changes that occur when you're running a machine.

Being able to identify that this is a problem, this is not normal, we're going to have an issue if we don't address it — that way, you can address the issue before it becomes a real problem — before the machine goes down. You can also work it into your production schedule more effectively. So, predictive maintenance is probably right now, one of the biggest uses of AI in manufacturing, and it's a valid use case.

Second area is just analyzing and summarizing vast amounts of data to aid in decision-making. Looking at all of your quality issues across the last year and looking for trends and exceptions and that sort of thing, I mean, it's not something that a human couldn’t do, but it takes a lot of time. With artificial intelligence and machine learning, those systems can go through that data very, very quickly and identify trends and exceptions and so forth. We have a number of solutions that do that. It's not just looking at quality — that's one example — but that ability to analyze and summarize vast amounts of data really helps with problem solving and decision making.

Number three, augmenting human skills. That's where our augmented reality and virtual reality solutions can come into play. For example, our augmented reality solution called DELMIA Augmented Experience actually learns part shapes. It takes a 3D model and learns the shape of a part. Now, how does that help? Well, it helps in a couple of different ways, because we support multiple different technologies, and I'm going to step aside for just a second to talk about that.

Most people, when they think about augmented reality, they think about HoloLens, Google Glass — you know, something wearable. And we support that. But we also support digital cameras in a tablet or a projection system for a large surface. This is really applicable for industrial equipment, or a large surface — we can project work instructions on the surface.

Now that projection system — I've seen other companies that can do that — but you have to be very careful on how you line that projector up with those systems. Because we have artificial intelligence built into ours, it learns the shape of the part. So, when you point the camera and project it to the part, it automatically aligns, because it knows the shape of the part, so it knows where to project the instructions.

We've got one example. It's the interior of a plane body and there are 200 rivets that have to be put in. What they used to have to do is go through and chalk mark all of those — for what size and type of rivet and that sort of thing — and it took them a couple of hours to do all that chalk marking.

Let's go back to our transcription error standard. Remember, studies show 85% is the best accuracy anytime you're doing manual transcription. So, they're chalking off 200 different rivets, and they probably got about 30 of them that aren't quite right because of that inherent error. It's taking all that time, and they have to go through and put in the ribs, and then they have to go and clean off the chalk.

Now they project it. The projector again, has learned the shape of that plane body, and it projects the right rivet size, type and depth and so forth, in the right location 100% of the time. That artificial intelligence, where it learns the shape of the part, really helps in terms of work instructions and improving production, productivity and quality. 

Then, in that type of environment and additional areas where our augmented reality solution helps with artificial intelligence, it is just guiding operators through quality checks. Again, it knows the shape of the parts. It knows on an assembly, what parts go where, what the orientation should be, and so forth. Now, it can walk the operator through that process again, with a projector, with Google Glass, with a tablet — whatever technology you want to use. But it's also got the intelligence built in that it can do the quality check for you.

We've got one customer example, where they hold a tablet with a camera up to the part. Now, they're seeing the physical part. And on their tablet, they're seeing the physical part with all the different brackets that go on, laid out on top of the physical. So, they're seeing the augmented, the virtual and the physical together.

And all those brackets start out blue because they haven't been checked yet. Then it automatically, using artificial intelligence, checks every bracket. As it checks the bracket, they go from blue to green, blue to green, blue to green, blue to red. Red tells me there's a problem with that bracket. It's either the wrong bracket or the orientation is wrong. So, it's doing that quality check automatically based on artificial intelligence. That’s augmenting human skills via — in our case — VR and augmented reality solution. This really is one of the areas where AI and machine learning can help in manufacturing.

Other areas that we are beginning to work on, or that we're investigating or developing, are things like generative design of new products and processes — having artificial intelligence do some of the design work. Now, there's always going to be the need for human engineers and for their knowledge and know how, and for their creativity, but artificial intelligence can be used to take the designs we've done in the past and based on that — and based on artificial intelligence and machine learning — begin the design of new products much more quickly. So, it makes the engineers more efficient, more effective and gives them a better starting point.

Another area that directly affects sustainability is the analysis of things like machine performance, heat and water consumption and other environmental factors to improve our sustainability. To identify if this machine is running efficiently or using too much power, which affects the CO2 footprint. It can identify a hot spot in a plant where the heat is too high. Maybe we can adjust our HVAC system to not only adjust that down so it's more comfortable, but also to reduce our heat usage and our CO2 footprint. So, all sorts of ways that type of analysis can help to improve sustainability, and these are really kind of just the tip of the iceberg, because the more use cases we identify and implement, the more use cases we will find. But all those really work together to show how artificial intelligence and machine learning can really transform manufacturing in a powerful and positive way.
 

Mike Bradford: If I needed to give a single piece of advice to manufacturers? One piece of advice that I would leave you with to future-proof your operations is to start now. Start now by getting educated. Go to events, get involved in organizations, talk to your peers, educate yourself, and then begin integrating and digitalizing technologies.

You can start small, like we discussed earlier. Pick one point, if you're doing really well in product design, design engineering, look at how you can integrate your manufacturing engineering and communicate better and improve that process, and then go to execution. Or if your shop floor execution is great, but the link back to manufacturing engineering and design engineering is weak, start with the execution piece and how it works with process design or with planning. It doesn't matter where you start. Find a place where you've got a weakness you know you can fix, or a place where this place is strong but the link to other areas is not. Find those good places to start and start. So, there are multiple approaches. The key is to identify your most significant deficiencies and start.

To discover more about how you can deploy effective strategies to achieve operational excellence, please visit www.3ds.com and check out the industrial equipment page, where we have more content on manufacturing and operations.

Narrator: Disruptors Unleashed is produced by Dassault Systèmes. For more episodes, follow us on Apple Podcasts, Spotify, Deezer, or your nearest streaming platforms. To learn more about Dassault Systèmes, visit us at 3ds.com.