Amid vigorous growth in their industry, product lifecycle management (PLM) software developers are exploiting the cloud and machine learning to manage data and enhance the users’ experience. Industry leaders are also looking at blockchain, either to expand use of the distributed ledger in their software or to integrate it where it makes the most sense in their PLM applications.
The PLM market last year grew 9.4%, with 18 providers generating related revenue of $500 million, said Stan Przybylinski, VP of research at CIMdata Inc. Looking toward 2023, CIMdata sees the PLM market growing to nearly $72.4 billion at a combined annual growth rate of 8.6%.
“Siemens PLM Software’s acquisition of Mentor Graphics made them the only provider with offerings spanning mechanical, electrical/electronic, and software,” he said. “Because of that acquisition, CIMdata’s estimate has Siemens in the No. 1 overall position, topping Dassault Systèmes by less than $100 million.”
With the enhancements and consideration of blockchain for PLM, the application is being redefined in part to reflect fundamental advances in manufacturing.
“I think the big issue is PLM’s not a standalone application anymore,” said John Kelley, VP of product value chain strategy at Oracle. “It’s part of integrated supply chain and manufacturing, and it’s a key piece of the transition to Industry 4.0 and smart, connected devices.”
PLM has also taken on management of digital twins, big data and analytics to build useful applications like those for preventive maintenance, asset monitoring and production monitoring, he said.
Siemens PLM Software CTO Jim Rusk might agree with Kelley. Rusk said comprehensive PLM software offers a competitive advantage.
“We’ve seen a pretty significant progression of how products are developed,” he said. “Historically, mechanical was developed by the engineering group and then maybe you’d have electronics that got developed by another group and maybe there was some software. Those domain barriers in companies that are producing complex products can no longer stand. It’s harder for them to compete when they’re serializing these processes.”
Adding gasoline to machine learning
Vikram Vedantham, senior manager of business strategy at Autodesk, said his firm’s generative design technology, originally introduced last year and now integrated within its design-to-manufacturing platform, Fusion 360, has been a leader in the paradigm shift to bring intelligent automation technologies to the broad engineering market.
Generative design takes into account the definition of a problem and any constraints imposed by a user—including dimensions, materials, loading conditions—to generate multiple viable design solutions or answers to an engineering problem.
“We are keen on removing the bottlenecks that currently exist in how data flows from one person and process to another, by taking on the burden of translating that data through a seamless end-to-end digitized experience from design through manufacturing,” he said. “Because then we can start looking at options where we can selectively serve up responses that a customer prefers as opposed to exposing them to hundreds of outcomes.”
Autodesk has enhanced the Fusion 360 user experience by enabling data generated from early-stage ideation all the way through to manufacturing preparation to move seamlessly within the workflow as users toggle among “workspace” pages.
“We absolutely want to make sure that we take on the burden of manipulating the data within the workflow,” Vedantham said. “Our customer research and our engagement with the customer definitely show that’s an emerging and imperative need in the market.”
While Autodesk works on integrating machine learning into its generative design process, Oracle is leveraging the artificial intelligence for predictive analytics and quality management in its Oracle PLM Cloud application.
“An IoT-connected device may detect a vibration that is outside of the design parameters. That’s where leveraging predictive analytics and machine learning come in,” Kelley said. “First, I need to predict when that asset will fail or when the quality of the product produced on that machine falls outside of parameters. Second, I need the analytics to evaluate the work order scheduled for the piece of equipment or product line and pick the best time to issue a predictive maintenance work order to ensure we minimize the impact on my factory throughput. But I’m also going to feed that back into my quality application to determine if product changes are needed to prevent or reduce the chance of this problem from happening again.”
As companies move from product sales to products as a service, guaranteeing uptime to control the revenue stream is critical. Predictive maintenance and predictive asset management are critical for these new business models and are now part of PLM software.
A digitally connected PLM is no longer used by just engineering, either. It’s also used by supply chain, manufacturing, sales, support and other partners across the product value chain and, as such, role-specific user interface and dashboards are needed to ensure each user gets useful role-specific information needed to do his job.
“So, we’ve had to focus on presenting meaningful information based on roles,” Kelley said. “And that’s key, because you need to store a tremendous amount of information about the product in PLM to create that digital thread or enterprise product record you want.”
Oracle does that directly in the UI and in analytics embedded in the application.
“As you’re doing your job you can see analytics and dashboards that are very configurable for your use and your role,” he said. “So, an end user is going to see lower level transactions while a manager’s going to see a dashboard about product health.”
At Siemens, machine learning is used in the integrated NX software for design, simulation and manufacturing.
“We’re gathering data as the user is working with the product on a user’s machine,” Rusk said. “But then we’re putting some machine learning to work and saying that given the last several hours or longer here’s a pattern we’re recognizing. Based on what that user or a larger group of users has done in the past, we can predict what it is they need to do next in order to complete their task.”
There’s still a need for human engineering judgment and making decisions, he added. But artificial intelligence and machine learning can get to those decisions and serve up the information that the user needs much more quickly.
Rusk said making the user experience simpler is “front and center” for its server-based PLM software Teamcenter and Teamcenter’s browser-based entry platform, Active Workspace. Together they’re the primary access to the information of a product that can be presented textually or graphically. That’s because the complexity of its group of users has increased.
“Our users are producing some of the most complicated products in the world, whether you’re talking about automobiles or airplanes or consumer electronics,” he said. “Whatever it might be it requires a broader and broader audience.
“You can imagine that going beyond the traditional desktop environment it opens you up for access from mobile devices,” he said. “It opens it up for other parts of the operation, including sourcing, shop floor, and the engineering department. The idea there is to simplify and flatten the user experience so users can very quickly perform a search, grab the information they’re interested in and proceed to the next step.”
Where does blockchain make sense?
Also part of PLM is track and trace, the system to pinpoint every location of an item and other information.
Oracle this year began offering blockchain applications for intelligent track and trace, lot lineage and provenance (which helps with unique device identifier compliance), intelligent cold chain (documentation of proper refrigeration for perishables) and warranty and usage tracking.
“We consider IoT and the ability to incorporate blockchain as part of our overall footprint,” Kelley said.
Oracle is evaluating blockchain for future bills of material and change management collaboration with outsourced design and manufacturing partners.
Kelley said traditionally the OEM will manage a supplier collaboration by managing the BOM (bills of materials) and engineering changes in their PLM system and then ship off the information to the suppliers who would then pull it into their system. But to his company, collaboration makes blockchain apps a good fit for BOMs and engineering change orders, and Oracle is building plans for them.
“So that’s an area we’ll be targeting for blockchain,” Kelley said. “We don’t have a date or time on that yet, but it’s definitely an area we’ll be leveraging with blockchain to increase the collaboration between partners and the product’s supply chain and value chain.”
Blockchain is an “area of exploration” for Autodesk, Vedantham said.
“We’re also looking at it from a strategic partnership aspect to figure out how best to integrate it,” he added.
And at Siemens, there’s no timeframe for implementing blockchain although it’s an “active area of interest” for the company, Rusk said.
“There’s a concern around counterfeit information that might have been picked up somewhere out on the digital space then gathered and delivered back to the OEM,” he said. “There’s potential for blockchain to be able to insert itself and prevent that from happening in those kinds of scenarios.
“So, I think this approach of allowing it to chase down particular concerns around proper validation, counterfeit, supplier validation, and other types of concerns you might run into in cyberspace is one of the key benefits you get from blockchain.”