PTC's pop-up Metaverse - no goggles required!

PTC's pop-up Metaverse

The original vision of the Metaverse spelled out a large-scale, fully interconnected, persistent virtual world. Sounds great until you have to build the whole thing out. PTC is exploring an incremental approach for building pop-up metaverses rendered on top of the physical world built out one service call at a time.

A service tech can pull out an iPhone to capture a digital representation of a part, assembly, factory, or service center with all its flaws and changes rather than the pristine version it was born. On the back end, this plugs into a spatial computing layer laying the foundation for new spatial apps, AI processing and enterprise apps like CRM, ERP, MES, ALM, and PLM.  Set the phone on the wall, and it can capture their skeleton (for privacy) at 25 frames per second to analyze how they work for ergonomics, coaching and billing purposes.

Vestas is using the tech to help teams collaborate on pop-up factories for building wind turbines. Burckhardt Compression is digitizing large compressors installed in remote locations to improve collaboration.

Mercedes Benz uses it to analyze ergonomics and maintenance tasks to train service techs. In Germany, car service technicians get paid based on how long jobs like a tire change are expected to take. If they work faster, they get paid more per hour. It also allows a single trainer to train remotely rather than driving or flying around to watch in person.

Product-centric view

PTC specializes in helping companies building complicated products with a long life span. Many live in the field for a decade or more. So, they have a product-centric view of the world. Kevin Wrenn, Chief Product Officer for PTC, argues that digitizing representations of these products is essential for digital transformation:

You have to start with digitally transforming your product, that has to come first. And so, if you have that mindset, product data becomes extremely important to digital transformation. And so you need to think about everything you do, from requirements, system models, and all the data you'd put into a PLM system. This includes CAD models, engineering bills of materials, manufacturing bills of materials, and service bill of materials that all become important data to drive the digital transformation.

Closing the loop

Equally important, enterprises also need to consider how to tie this together with back-end data scattered across various IT systems. PTC has built up a portfolio of engineering, manufacturing, and service solutions, which Wrenn believes is crucial for completing the circuit in closed-loop lifecycle management. PTC acquired ServiceMax last January to expand its reach into the service industry. It also acquired OnShape in 2019 to flesh out its cloud strategy for collaborative engineering and design.

Wrenn says:

A lot of important product data is in the factory's ERP, MES, and quality systems. But then, if you think about the product that's been living out in the field for decades, service information becomes really important. Because if you want to do closed loop lifecycle management, and you don't have a big position in service, for those kinds of companies, you're missing a system of record for the longest part of its life.

A spatial computing layer

Valentin Heun, VP of Innovation at PTC, sees this work as part of the evolution of the computer interface toward spatial computing. He says:

Look at the history of computing. First, we have text, then the desktop paradigm, but we got a little bit stuck there. Now, we need spatial computing, which is the idea that the computer is not on a desktop but around you. A car is a good example of what spatial computing can really look like.  When you drive a car, you actually don't drive a car directly. You have a joystick in your hand that operates a computer, and the computer drives your car. The car drives by wire, so you are actually using spatial computing when you drive a car.

Heun envisions a world with a spatial computing operating environment that runs spatial applications networked among multiple users over the Internet. He started work on an open-source project at MIT a few years ago called the Reality Editor, designed to provide an IoT platform for physical spaces. At PTC, he helped modernize this into a spatial toolbox for running applications.

He sees existing approaches like using USD and glTF for sharing digital twin data as a content layer for the metaverse. Kuhn is excited about building another abstraction layer for running new apps on top: 

What is really interesting for us is creating an application layer and an operating environment layer. Once you have control of how content is processed or how you can deploy logic into space, you can load the application from another party or bring all the PTC content into one 3D environment. And the way that you can create that application layer is you have to actually build a system around it. You have to build a garden where you can put the plants or a box in which you can put all your tools.

Generating requirements

Like all companies these days, PTC is also getting on the generative AI bandwagon, but for designing products rather than text or images. Engineers can set up the space, connections and boundary conditions and automatically generate a part optimized for a particular manufacturing technique. They can also generate an array of variations using different manufacturing processes to provide various options.

Down the road, Wrenn envisions extending generative AI to help engineers complete requirements like GitHub Copilot can help complete code:

You can imagine a ChatGPT kind of engine saying, ‘I have seen requirements like this before, let me finish them for you.’ Then, with that set of requirements, let's search your PLM database and see if a part already fulfills those requirements. If not, it could funnel those into generative design to automatically generate a part.

My take

Like all PLM vendors, including competitors Dassault Systèmes and Siemens Digital Industries, PTC is building an onramp to its own industrial metaverse ecosystem for creating digital twins. For now, at least, these ecosystems of cloud services and interconnected digital twins and digital threads all operate separated by moats for various technical and business reasons. But regardless of the platform companies that build things take, digitalizing physical things is key to enabling the kind of rapid innovation of digital natives like Uber, Airbnb, and Google.

The challenge is building an on-ramp to help enterprises digitalize their existing products. A pop-up metaverse generated on the fly by an iPhone seems to provide an easy way for getting started. Rather than requiring a big bang, enterprises can start small just to improve their service process and then build out the other kinds of capabilities as they need them.

Closed-loop lifecycle management will become increasingly important as enterprises find ways to transition to a more sustainable circular economy. Better digital twins that capture how things break and why will make it easier to improve the reliability and cost-effectiveness of products. It will also allow companies to analyze the total cost required to support new XaaS (everything-as-a-service) business models.

The privacy-preserving human activity tracking is pretty essential too. This kind of approach will make it easier to identify how people perform tasks to think about manufacturing or repair process redesign in the same way, companies use process mining today. Now if they could just come up with an app to improve my golf swing…