originally published in Plant Magazine
Theme park executives and ride designers watch as their newest rollercoaster crests the first hill. The drop that follows thrusts the coaster through the loops, turns and other thrills that will lure thousands of adrenaline seekers a year or two from now.
Full disclosure: nobody rode the coaster. It doesn’t physically exist. The entire “ride” took place on the servers of Port Coquitlan, BC-based Empire Dynamic Structures Ltd.
What the “riders” watched was nothing new to fans of today’s video gaming consoles. The remarkable aspect of this visualization is the process used to develop it—an emerging design-build-automate paradigm that has already made waves in industrial design ponds.
“For an amusement ride, we can make an animated model that shows what it looks like from the outside when you stand on the platform watching it go by, or what you see when you’re sitting in the vehicle,” explains Craig Breckenridge, Empire’s drawing office squad leader.
Use of digital models goes well beyond product demos. Cutting-edge use of advanced digital design tools (in this case, Autodesk Inc.’s Inventor), represents the widely acknowledged future of manufacturing product lifecycle management (PLM).
According to Dick Slansky, senior analyst and PLM research director of the ARC Advisory Group based in Boston, PLM includes every step in the existence of a given product, from concept and design through manufacturing, commissioning, rollout and support, right to obsolescence.
To build a full, working prototype of a rollercoaster prior to the final product would be prohibitively expensive. But in a perfect world, rollercoasters should only throw twists at the people who ride them, not the people who design and build them. That’s why manufacturers increasingly rely on the latest PLM technologies.
To make today’s PLM tools work, manufacturers and suppliers shift their focus away from drawings to parameters. Industrial designers, engineers, shop floor manufacturers, purchasers and even customers contribute information to a central model and track how the work of others affects their own.
Consider the radical changes aircraft manufacturer Boeing went through. Boeing’s case is one of many examples Anthony Williams documented in Wikinomics: How Mass Collaboration Changes Everything, which he co-authored with Don Tapscott.
Boeing realized it could not, on its own, take entire aircraft from idea to market quickly enough to stay competitive, so it became a “prime assembler,” largely by divesting itself of non-core operations.
Suppliers were asked to design major portions of its 787 Dreamliner aircraft, an acknowledgement by Boeing that the suppliers knew specific systems better than it did.
Radical software and collaboration methods don’t just help the technology professionals. “Our sales guys are not engineers,” acknowledged Matko Papic, manager, engineering and product development for Evans Consoles Corp. So leveraging the parts commonality and logic that Evans adheres to when it designs control consoles, the Calgary-based manufacturer created user-friendly Snap software to help sales representatives create designs when they meet with customers.
“It contains all the product management controls that we want,” says Papic.
Snap itself is an interface that ties to systems such as Autodesk Inventor and Microsoft Axapta, the enterprise resource planning (ERP) system Evans uses. These “back-end” systems check inventory for parts and perform other preparatory tasks. Meanwhile “whoever deals with the customer can instantly generate a full proposal package—all drawings, price quotes, pro forma invoice, renderings and so on,” says Papic.
In the next phase, the tie-in to the ERP will allow Evans to bypass both the project management and design group for proposals that don’t require extra customization. “We can send these solutions directly to the shop floor.” This set of innovations pares the cycle time from initial sales visit to shipment of the final product to two weeks from a pre-Snap four weeks.
Slansky is bullish on virtual commissioning, particularly for the automotive industry that, he opines, must create new products for niche markets more quickly than ever and in smaller quantities, and therefore must more quickly commission changed production lines as well as the new products they create.
“It used to take months for groups of engineers to physically commission stuff and make it work,” he says. “Now all that is gone.”
Empire’s design and engineering groups rely on Inventor to track key parameters for them, enabling individual parties to raise issues at the moment they appear in the digital model.
For his part, Breckenridge is thankful he no longer has to count parts. “It’s easier to purchase 500,000 bolts in time for production when you know you need 500,000 bolts, and what size you need, months in advance,” he says.
Yet despite all the intellectually appealing underpinnings of digital prototyping, promise still exceeds reality.
Prototyping limitations
“There are things that digital prototyping won’t tell you,” Breckenridge says. “It doesn’t tell you how quickly something will wear. When we put a vehicle on the track, the wheels will wear—but how quickly do they wear? You can calculate to death how much the wheels deflect, but how much do they really deflect?”
“Anything that involves passengers, we need to be extremely careful with.”
Papic concurs. “When a customer has specific requirements, we would rather over engineer a solution and be safe than spend time testing it. We can do that in general platform design, when we launch a product, but when we respond to quick requirements, we over engineer for safety reasons.”
Empire also builds large telescopes. For a 30-metre apparatus in Chile that should be complete around 2018 (according to the website tmt.org), Empire will not rely on digital models to calculate the time to assemble the structure.
“We’re going to build a physical model of a section of the telescope enclosure and time it,” Breckenridge says. “We’re going to compare our estimate based on our digital prototype and extrapolate how long it takes when we build the actual enclosure to determine total construction time.”
On top of the shortcomings of today’s tools, none of them cover the full range of Slansky’s cradle-to-grave idea of PLM.
Even if such a system existed, long-entrenched “siloization” among suppliers can hinder its use. Boeing requires all suppliers to collaborate if they are to contribute to its planes, but not all prime assemblers have that kind of pull within their supplier networks.
Such difficulties occur within companies too. “Our customers are starting with engineering, making sure electrical and mechanical engineers are tied together,” says Kerry Samurs, sales development director of Autodesk’s Americas manufacturing solutions division.
Subsequent integration steps are analysis, then ideation, which has typically worked separately from engineering.
“Often it’s because they’ve been using two different tool sets, or the creativity side is done on paper,” says Samurs.
Not all the change is technological. Williams noted older workers, who came of age in more rigid, hierarchical environments “tend to be less comfortable with horizontal and fluid ways of collaborating.”
Their bosses, meanwhile, contend with the notion that competitors can also be partners. That opens up a debate over proprietary versus shared intellectual property (IP). The “wiki workplace” demands greater sharing of information, and as companies venture deeper into collaboration, they agonize over “open sourcing” IP, something that historically is not done.
Boeing got around this issue by producing the 787’s tail fin in-house, which suggests other aspiring prime assemblers look at what elements of their IP truly sets them apart, then decide whether and how to protect it.
Easy answers? They’re as easy to come by as a precise count of bolts in an amusement park, particularly since technology continues to stride ahead of many manufacturers’ ability to keep pace. But in Slansky’s view, they don’t have much choice. “Companies must innovate or die,” he shrugged. “It’s part of the business.”