New production materials and methods have begun to revolutionize how we make things. If we look at individual manufacturing advances in areas like the use of carbon fiber in passenger aircraft, advances in Computer-Aided Design and Manufacturing (CAD-CAM), additive manufacturing (AM), or industrial robots streamlining work on factory floors, there are countless ways that advanced manufacturing is helping us to make things lighter, cheaper, and better.
But how do you tie all the potential opportunities together to reach the optimum solutions? That’s where the concept of generative design can help bring all the different facets of modern production to bear on producing the best answers possible.
“The challenge we see in manufacturing is that people don’t just want the iPhone – they want the latest iPhone sooner,” said Scott Reese, SVP of Manufacturing, Cloud and Production Products at Autodesk, a software provider for 3-D design, engineering and entertainment. “That leads to lots of questions about how to get there. How does additive manufacturing fit? How do cobots fit?”
The endless challenge of meeting consumer needs coupled with today’s drive to reduce resource usage and decrease environmental footprints brings enormous pressures to bear on manufacturers.
Fortunately, modern technology offers solutions. “Generative design brings AI to the design process,” explained Reese. “Take the example of making a coffee cup. In the past, someone would bring a problem or an idea to designers. They’d start with a CAD drawing, work up some different concepts, send those for testing, finalize the design, and send it to manufacturing to produce. But the best designers – smart, well-educated – can still start with a bad idea.”
Generative design backs things up a step. “Instead you start by describing all the parameters involved: it has to hold a hot liquid; it has to cost less than $3; it has to be comfortable for a person to hold; and so on. Generative design turns it all into equations, and will generate thousands of options, then identify the best one. For its solution, the outcome has already been validated to fit all the user requirements, and has also been modeled for the manufacturing process.”
As it’s freed from human preconceptions, generative design can result in unusual solutions. “When we used it to design a chair, what it came up with looked nothing like a chair you or I would design,” Reese said.
Airbus is a customer that’s already benefited from generative design. Three years ago they did a proof-of-concept redesign of an aircraft partition, which was originally a solid wall. The design that was generated – the “bionic partition” – was 45% lighter, as it was a lattice rather than solid, but it was just as strong as the original partition. Just by rolling the design out in its backlog of A320 jetliners, Airbus calculated, could eliminate about half a million metric tons of CO2 emissions.
However, that design relied on metal AM technology for fabrication, which wasn’t feasible. A later, more mature version of generative design was able to optimize production for other manufacturing methods. For “bionic partition 2.0,” the optimum solution was to design a mold for the new part, then cast it in an alloy that was already approved for aviation applications. Airbus hopes to complete its first prototypes of the new partition by the end of the year. Meanwhile, they’ve begun looking at a variety of its structural aircraft parts to see if generative design can help make them lighter as well. They’re also using the technology to help redesign their factories to be more logistically streamlined and environmentally friendly.
General Motors (GM) has also made use of the technology, in their case to reduce complexity. They set out to improve a part almost nobody ever thinks about – a seat belt bracket. “Located under the seat, it was made of eight separate parts,” Reese said. “Each of them came from a different supplier, meaning an equally complex supply chain.” GM plugged their parameters into the software, including the supporting car floor design, design loads, and size constraints. They wound up with a new one-piece part that was 40% lighter and 20% stronger than the original.
Even with complex designs, though, maturing advanced technologies will further improve outcomes. “As AM comes online, complexity becomes free,” said Reese. “The additive process doesn’t care how complex the design is.” A good example of that is the generative design for an improved crimper used by electrical line workers. The Breakthrough Innovation team at Stanley Black & Decker wanted to reduce the weight of the 15-pound tool, which is held overhead in one hand by the worker. As it has to handle a 15-ton force, the new design had to be at least as strong as the original. The team was able to evaluate the top ten of 100 ideas generated, and chose a design that cut over three pounds from the tool, and could be made via AM in 20 hours, while the original tool took over a month to fabricate, requiring five separate operations spread between different forming machines.
The potential applications will only increase over time. “Secular trends like the electrification of vehicles and the Internet of Things – think about where your refrigerator knows you need milk – will make it harder and harder for manufacturers to keep up,” Reese explained. “We’re starting to see disruption in manufacturing because of that. The way people think about manufacturing today won’t work in the future. We can see production moving much closer to the consumer, with factories everywhere. You could shop at an online retailer, pick a customized product, and the design could be sent to a factory close by you for fabrication, with almost immediate delivery. AM and other advanced manufacturing techniques help make that possible, but by themselves they won’t get us there. We see generative design tying it all together.”
Originally published at Forbes.com on November 25, 2019.