Intelligent systems engineering with modelling, simulation and AI

How integrating modelling, simulation and AI is paving the way for automated product development and intelligent systems engineering.

Sustainability is becoming more and more embedded into product design processes across a multitude of industries, from cleaner vehicles and more efficient industrial equipment to advanced materials engineering. A combination of modelling, simulation, artificial intelligence (AI) and machine learning (ML) is helping this evolution on its way in the form of autonomous mobile robots (AMRs). But while great strides are continually being made using these innovative technologies, there is still untapped potential waiting to be unleashed, believes Philippe Bartissol, VP of Industrial Equipment at Dassault Systemes.

“Product and machine design is evolving through what we call MODSIM AI, which is modelling, simulation and AI all together in one platform – 3DEXPERIENCE – also with sustainability calculations and Lifecycle Analysis (LCA) added in,” he says. “When you develop a new product, you should also think from a service engineering perspective: How am I going to service, maintain and retrofit the new range once it reaches the market? It is not only product design engineering that matters, but also the system, production and service engineering processes you offer throughout that product’s lifecycle.”

MODSIM CAPABILITIES

MODSIM unifies modelling and simulation on a common data model within the 3DEXPERIENCE platform to allow engineers to consider the entire product development process for an AMR in one place. More than just simulation-driven design, MODSIM enables simulation to drive the entire product development cycle from beginning to end, including requirements, validation, certification, development, programme management, design processes and automation.

“We have plugged the largest database of CO2 emissions calculations into the 3DEXPERIENCE platform, so that engineers can explore all aspects of the product design process,” Bartissol explains. “Engineers can simulate design and material options, manufacturability, modularity for disassembly later on, and for the calculations there is an LCA capability built into the platform that, when engineers are generating design, material and manufacturing alternatives, the system will tell them immediately what the CO2 impacts will be. Then, from a sustainability perspective, engineers must look at longer life pieces of equipment and future product lines.”

This is where designing for retrofit comes into the conversation, he adds. “Putting together the business case of retrofitting versus new, engineers need to consider how much value they can increase if they retrofit, and what is the cost. Looking at this like an infinity loop, you would have in the beginning design engineering, manufacturing, service engineering then selling, producing and installing. Then, during the life of the equipment, there will be a value for certain costs. When the value is not enough in operation while using the piece of equipment, engineers have two possibilities: either to retrofit or to buy a new product. So, by working on the after sales of the retrofit possibilities, engineers can extend the life of a machine or product with less of an impact on the environment.”

MODULARITY FEEDS CASE FOR RETROFIT

To enable greater retrofit possibilities, designing AMRs and other machine equipment for modularity from the start is crucial, Bartissol says.

“We have all these initiatives to design for cost, for service, for simulation driven design, but this is  not enought. We should be designing for retrofit and that should be the primary goal for design right now,” he explains. “Engineers need to take the view: what will this piece of equipment become over the next 10, 20, or 50 years? This is important on two counts, for the environment and sustainability but also for profitability.”

A modular product range demands advanced design software capabilities, such as modelling and simulation. “You need to have a software or PLM simulation platform that sustains modularity,” Bartissol adds. “With this, you can carry out configuration, define modules, interfaces and so on. This is what we are striving for. With these capabilities, you can choose your reason for retrofitting: to consume less energy, to save water, to reduce noise, to be more agile, or to be IoT capable, for instance.”

AUTOMATION IS THE ANSWER

“Due to the current skills and worker shortages across production industries, we will see more and more automation,” Bartissol predicts. “We used to have very long runs in production lines but now these have shortened, meaning we need to work in a more agile mode to rapidly adapt to changes. We now see a new domain emerging called intralogistics, which is the merging of robots, AMRs, forklifts, conveyors and automated storage capabilities, among other things. All these elements are connected by a global controller PRC or MES to create wholly automated warehouse systems.”

Looking to the near future, Bartissol foresees the combination of AMRs and AI to be adopted effectively in logistics centres and production lines, with humans playing their part at the beginning and end of the product development and manufacturing process, but not in the middle.

“Most of this will be robots working together with human beings in a safe and optimised system,” he says. “The actors are transforming themselves, and we see more and more integration capabilities and layers all in one place. The ones who are investing now and investing heavily are the ones who will likely win in the future.”

DIGITAL TWIN SIMULATIONS

As Bartissol says, simulation-driven design is closely interlinked with digital twins, where engineers can model and simulate various scenarios and designs of machines in order to capture and feed data into the AI algorithms on the 3DEXPERIECE platform.

“Engineers can adapt the digital twin of a machine to optimise different simulations and analyse the data, such as the event of a machine crash or failure,” he explains. “Through this, you can see the cause and understand how to prevent this. Also, by observing the behaviour of the equipment over time, you will be able to predict the next failure. Simulation of models helps you to not only understand the past, but also navigate into the future and improve overall equipment efficiency.

“Now, some of our customers are asking us to create specific twins of each machine: an engineering twin, a manufacturing twin, and in the future a field twin. This allows the opportunity to switch to an equipment-as-a-service business model. MODSIM AI will serve not only during new product development, but also when the equipment is in operation and requires maintenance.”