Avi Reichental is CEO of Techniplas, a supplier to companies such as BMW, Daimler, Ford, and Fiat. The company has developed a solution using 3D printing and AI to make parts up to 47 percent lighter. Using AI technology, Techniplas is able to automatically calculate how to minimize the amount of material used to create parts that are just as strong as traditionally produced and heavier parts.
With 3D printing, a part can be made faster and at lower to cost to be any shape instead of making elements that would need to be connected afterwards. Techniplas Prime, the world’s first and only one-stop shop for automotive e-manufacturing platform, makes the process even easier as manufacturers can simply send their 3D printed models to have parts 3D printed on spec and delivered.
Reichtental was interviewed by Senior Editor Bill Koenig via email.
How does your company combine AI and 3D printing?
Let me start by saying that Techniplas is a leading global manufacturer of automotive and industrial products with factories around the world and over a century of traditional manufacturing expertise. Over the past couple of years, the company has begun an extensive digital transformation starting with implementation of industry 4.0 capabilities in operations and implementation of digital manufacturing and generative design capabilities primarily for the purpose of greater manufacturing flexibility and lightweighting capabilities. The company also has a unique cloud manufacturing platform called Techniplas Prime and it already uses this ecosystem to add capacity regionally around the world in order to scale and localize global delivery.
In our case AI helps us do a few things at the design phase primarily using algorithmic topology. Working with our open innovation partners in generative design and additive manufacturing, we believe there is only one correct answer for each design optimization problem, unlike other suppliers that could offer you several dozen solutions, with the engineer choosing the final one. We consider that if you set the design problem correctly, there is only one accurate answer and AI helps in determining that. Secondly, the only way to compress design time and cost and help our customers get to market faster is through a design that is actually manufacturable, not one that requires processing or finishing by a designer or engineer. So, we ensure that we generate fully watertight and 3D printable files that are ready to go.
We are now developing additional capabilities to optimally orient and support parts in certain print engines as well. In all these areas, AI is helping to automate and optimize the outcome. We end up with a single design that is readily manufacturable and available to 3D print, injection mold or investment cast.
The unique aspect about Techniplas is that its optimization and lightweighting, powered by our open innovation partner ParaMatters is completely autonomous and AI enabled. It runs completely automatically and defines desired lightweighting within constraints you set. It runs in the cloud and for a typical size problem you will get an answer that is ready to be 3D printed or investment casted in about 15-20 mins without any human intervention. Once you get the desired solution, you can decide if you want to 3D print or if you want to bring it back into your own CAD in a parametric environment and continue your design.
Are there certain types of components where this technology applies more than others? Or is it applicable throughout a vehicle?
We can divide components into a few sections. There is high interest in under-the-hood high temperature components, and cabin parts, especially when you consider that steering wheels will disappear with self-driving cars and the function of the cabin will change. This will open up new experiences and use cases within the vehicle. Additive manufacturing can also be used to develop parts for the undercarriage such as brake drums and trailing arms. For interior components, we’re targeting dashboard inserts, cup holders, storage components and work surfaces within driverless cars as the utility of surfaces within the interior are shifting to productivity and entertainment.
In terms of materials, does your technology apply more to ultra-strength steel? Or to aluminum? Or to a variety of materials?
Direct metal 3D printing is not only the hottest growing segment over the past 6-7 years, it’s going so well that there are now more 3D direct metal printers installed than all other 3D printers combined. So, it has certainly taken the market by storm. This is because you can print final parts and products from engineered metals including ferrous and non-ferrous alloys. The utility is very significant, the productivity levels are acceptable, and the ROI is attractive because at the end what you get is a very simplified and condensed manufacturing process that gives you a net shape or near-net shape in a density and performance characteristics that is well understood. More importantly, it allows you to take advantage of new capabilities like generative design and topology optimization, which allows you to substantially lightweight products and, in many instances, combine assemblies. This is a key driver in auto manufacturing today because automakers follow very strict guidelines to take weight out of the vehicle for energy efficiency reasons.
With additive manufacturing, complexity is free, which means you can mimic the kinds of efficient high performance structures we find in nature – structures that were patented millions of years ago. Think of how efficient your bone structure is for example. Now, after millions of years, with the help of generative design, AI and topology optimization, we can crack the code of these incredibly efficient lightweight structures and reproduce what nature patented millions of years ago. That’s now being applied in the auto industry and in aerospace under the masthead of lightweighting. And what happens if we can take a dozen smaller parts and combine them into a unified assembly eliminating costly assembly labor and simplifying the entire supply chain? You eliminate the sub-assembly line, you eliminate the extra parts and the assembly fixtures required and you end up with a more efficient integrated part.
This is why we’re seeing an explosion in the adoption of metal printers, especially in aerospace, automotive, and also in the medical device industry – implants, knee, hip, spine implantable devices, etc.
Is this technology being used for parts in production vehicles now? If not right now, how soon?
The fact is that the automotive industry was among the first to adopt Additive Manufacturing (AM). As CEO of 3D Systems back in 2003, some of my first customers were Ford, BMW and Chrysler. These companies were all early adopters and used AM for design and rapid prototyping. Even back then, leading automotive and aerospace companies understood that this technology was a game changer. Fast forward to the here and now they each own and operate hundreds of printers for their design and manufacturing workflow. Material performance and scalability were the main things holding them back, but now the processes are more cost effective, more scalable, and with shorter development cycles. It’s a perfect environment for adoption of 3D printing.
How long did it take to develop the technology? When was it first devised and when was it presented to automakers/customers?
Every so-called exponential technology takes a few decades before it makes its breakthrough into the mainstream. So if you look for example at technologies like robotics that are all the rage now, those started in the early 50’s out of MIT.
If you look at 3D printing, the original concept was disclosed in 1980 by a Japanese researcher, Dr. Hideo Kodama, who came up with the first idea of layer-by-layer geometry assembling or manufacturing. Then in the early 80’s, between 1984-86, my colleague Chuck Hull, who founded 3D Systems, filed his first patent on the stereolithography apparatus. And, of course, Chuck went on to establish 3D systems; where I was privileged to serve as President and CEO for 12 years.
Then, there was another very prominent inventor in the field named Scott Crump, who is credited with inventing another 3D printing technology called FDM (Fused Deposition Modeling (FDM), or Fused Filament Fabrication (FFF) Scott started his company in 1988. So, if you look at it from 1980, the concept was already in active commercial use, as is the case with most nascent technologies. The supporting infrastructure, like ubiquitous connectivity, or the ability to do an infinite number of computational transactions in the cloud, the need for good integration with design tools, the whole workflow, or what people call today the digital thread, wasn’t even conceived of. So, to a large extent, the early 3D printers were a lot more like a Ford Model T rolling off the production factory in Detroit before the highway system was developed. There was no infrastructure for it. Also, like the Model T, the early 3D printers were very basic with very limited capabilities.
What we’re understanding more and more, is that for fundamental tech disruption to happen — if you look in the field of AI or machine learning, if you look at the field of IoT or sensoring, if you look in the field of 3D printing – it’s not enough for the field itself to evolve and become capable or affordable. What’s required is a convergence of enabling exponential technologies to allow the fundamental technology to mainstream.
So, in the case of AI, what was missing for decades was infinite and affordable computing power in the cloud. And, of course, the availability of data that could be mined and tagged and categorized so that all these amazing algorithms could be applied at scale. In the field of robotics, what was missing was better computational power and connectivity, affordable and diversified sensoring, and so on.The same applies in 3D printing. What it needed to evolve was more computational power and ubiquitous connectivity. It needed the parallel universe of design software to evolve, and the scanning software to evolve. It needed the chemistry of digital materials to evolve and for thermoplastics to evolve sufficiently. We needed the overall enabling infrastructure that connects it all in order to make it an accessible utility to industry, art, education, architecture and so on.
So, it typically takes all these ingredients and all these enabling technologies to converge in order for a fundamental technology like 3D printing to mainstream and reach its full potential and it’s happening as we speak.