You have heard it before, today’s manufactured products are becoming ever more complicated. As computers and microcontrollers get ever cheaper and more powerful they have become more enticing for product engineers to use and incorporate. This means the intellectual property in the embedded software has grown increasingly in value – possibly exponentially. This is true of practically every manufactured product today. 100s of millions of lines of code exist in airplanes, possibly 100’s of thousands in refrigerators and coffee makers. Automakers in particular are struggling with the transition of so much value existing in code, computers, and systems engineering expertise in addition to ‘hard components’ such as engines and wheels.
Complicating their life even more is the continuing rapid evolution of products, especially light vehicles. Within a few years, we could expect to see partially electrified, semi-autonomous, connected vehicles dominating new vehicle purchases. In smaller numbers, will be fully electrified, fully autonomous ones. Warranty issues and recall risks will only increase exponentially.
The field of Product Lifecycle Management, or PLM, was created to help manufacturers cope with the myriad elements of a product from inception to graveyard. Most PLM systems used today are extensions of Product Data Management systems that were originally used to manage CAD data. They are extensive, popular, and many organizations have wrapped their product development processes around them.
Given their roots in managing CAD, are PLM systems evolving to manage this greater emphasis on systems and software? According to a report by CIMdata, maybe not. In a commentary published on the consulting firm’s website in May 2016 “few PLM implementations achieve their original goals and often end up as PDM silos that struggle to evolve as business needs change.”
Enter a smaller player in the field with some radical notions about PLM software. The two ideas I think the community of users and competitors alike need to be exposed to are Flexible and Free. I explored these topics in a discussion I had with the CEO of Aras, Peter Schroer. His company offers a different kind of PLM in a radically different kind of business model.
Flexible – Addressing the Business of Engineering
“Up to about 10 years ago, most organizations wanted to manage their CAD better and maybe they wanted an Engineering Change Order, or ECO, process,” explained Schroer. That changed and quickly. “Suddenly we had people showing up wanting to manage manufacturing work instructions, technical documentation, simulation data, and supply chain issues,” he said. “We realized that this was developing into a really big picture.” That big picture he condensed into a simple catch phrase – The Business of Engineering. “It was no longer simply about having the right materials going into a chassis or a 3D CAD tool for rendering a surface, but the problems that seemed to be surfacing were about this Business of making products.”
That is where it gets interesting, as they say, from a data model point of view. The data model at the core of their system is not fixed. All data management systems have a data model at their core, detailing how different pieces of information affect the overall picture of a product. For example, a CAD model describes the physical shape. Add Product and Manufacturing Information, such as GD&T, and the data model now describes how to make the product. Attach material cost information and a manufacturer knows how to price it without losing money. Software algorithms, systems engineering math models, and other functional or behavioral models explains how it is supposed to work. And so on.
“We have a data model that is fluid and changes with the requirements of the customer,” explained Schroer. While a hard coded data model would need the user (or provider) to know what data elements are needed for a product, according to Schroer the Aras Innovator model recreates this on the fly. “It is based on XML, does not require hard coded relational data models, and can adapt to the changing business needs of the engineering organization,” he said. He further claims that the dynamic data model is naturally more efficient for big models. “It is much more scalable than a hard coded model,” he stated.
I asked him if he had a single phrase of advice to offer. “Embrace the bigger scope of the Business of Engineering,” he replied. “We can run data models with 10’s of thousands of elements. The boundaries and limitations of technology have gone away. It sounds scarier than it is.”
Free Open Source PLM
On top of that, an initial installation of Arras Innovator software is free. Yep, free. It is a business model that combines Open Source with a Software-as-a-Service subscription, or SaaS. The SaaS model is becoming more common in consumer oriented software. For example, I now pay a nominal fee per month for my Microsoft Office subscription. OnShape, a new CAD company that offers its software only through web-browsers (I wrote about this last month) is another in the field of engineering. “Combining Open Source with a SaaS Subscription business model is truly a unique approach within the PLM industry,” claimed Schroer. “The open download allows companies to start their PLM journey quickly, cost effectively, and on their own schedule. When they are ready to deploy Aras PLM into production, they have a choice: continue using the software for free or to purchase a Subscription for full turnkey support.”
He points to RedHat and its LINUX software as a comparable business model, with an open source platform at the core of a money-making business enterprise. He describes Aras “providing stability, security and features that make the technology safe for a mission-critical product design and manufacturing data” beyond the open source version. If a company wants that, or wants Aras to provide it. Otherwise, they have access to an open source PLM to do with as they please.
While free contrasts with the extensive licensing model most suppliers offer, there is another dimension where Aras differentiates itself. According to Schroer and experts like CIMdata it is unusually flexible. “Considering the accelerating rate of change in vehicle technology, and the increasing design and manufacturing challenges we are all facing, a little flexibility may be exactly what our industry needs for the next 5-10 years,” said Schroer.