Apple’s main manufacturing partner, Foxconn, has announced it is working with Nvidia to build digital twins that it says will reshape the future of manufacturing and supply chain management.
Nvidia and Foxconn last year announced plans to use Nvidia’s Omniverse platform to create 3D digital twin tech with which to plan and simulate automated production lines. The scheme was first put into effect at Foxconn’s Hsinchu factory in Taiwan and will be scaled out to Foxconn factories worldwide.
What happens in Hsinchu…
Apple’s connection with the Hsinchu facility isn’t particularly overt, but it certainly exists. There’s an Apple Store currently hiring in the city, and Apple also has an R&D facility there. In 2020, Apple confirmed plans to build a new plant in Hsinchu Science Park to supplement the operations it already had in place.
As far as we know, Apple’s Hsinchu-based R&D teams are working on next-generation monitor technologies such as low-temperature polysilicon displays and metal-oxide-semiconductor screens, along with quantum film image sensors, according to earlier reports. (Who knows, it’s not impossible that new tech used in the latest MacBook Pro displays might have been developed there.)
While a bit of a long shot, some of Apple’s server development team might also be based there, given the company is developing its own servers to support its Private Cloud Compute systems for Apple Intelligence. It was recently reported that Apple has asked Foxconn to make AI servers based on Apple Silicon in Taiwan, and given the proximity of the Hsinchu digital twins project, it is hard to ignore the overt opportunity for additional cooperation between the firms.
When it comes to manufacturing, Apple has a pressing challenge to scale up the capacity to build iPhones at factories outside China. Some of this work is already taking place in India where the company is rapidly ramping up production, but it is possible Apple wants some manufacturing taking place elsewhere, such as in Mexico.
Foxconn’s move to build heavily automated production facilities could help Apple with those efforts.
Industry 4.0 and the Apple supply chain
I see the latest news with Nvidia as part of a continuum. Foxconn has already built a growing network of eight Industry 4.0 lights-out factories in Taiwan, China, and Mexico. In China, the steady move toward additional automation means Foxconn has been able to reduce its workforce by more than a third while maintaining production levels.
Foxconn’s entire Shenzhen, China, Guanlan factory operates without lighting as it is automated and controlled by a cloud-based AI. The vision of that latter project is that it will become possible to simply tell the cloud AI what products need to be made and how they are to be manufactured, and the system will adjust itself to automatically churn those products out.
There’s a ways to go before that becomes possible, but it sounds like Foxconn will use Nvidia’s tools to track existing manufacturing processes so they can be more easily replicated at factories situated elsewhere.
“Through this technology, Foxconn can replicate and establish production lines across diverse geographical locations with unprecedented speed and precision,” the company said. “This capability enables Foxconn to swiftly deploy high-quality production facilities with unified standards in strategic markets worldwide, significantly enhancing the company’s competitiveness and adaptability in the global landscape.”
Digital twin tech is also very good at identifying bottlenecks and inefficiencies in existing production processes, while the ability to more easily take manufacturing lines to new nations also help build resilience into manufacturing systems. “When facing supply chain disruptions or sudden market demands, Foxconn can quickly simulate manufacturing process changes and adjust production strategies to flexibly allocate resources across different regions for itself and its clients, ensuring production continuity and stability,” Foxconn says.
Resilience and flexibility
To some extent, the writing has always been on the wall. Supply chains globally buckled during the height of the CoVID-19 pandemic, and Apple’s decision to widen its manufacturing base to new nations was a direct response to this. Apple — and quite clearly, Foxconn — now understand the need to build resilience into the supply chain, and one way to do that is to turn to using heavily automated manufacturing systems that can be easily set up and made productive in new locations. This seems to be the game in play here, particularly in the wake of Apple’s purchase of Darwin AI earlier this year.
The other part of that game reflects the challenge of staffing manufacturing operations at the scale Apple demands. Hundreds of thousands of people globally are now involved in building Apple hardware, and the job is skilled enough that recruiting all those workers can pose problems for the company. This is likely why in June it was revealed that Apple intends to replace 50% of iPhone related assembly line workers in the next few years. That ambition logically requires the kind of productivity enhancements Foxconn and Nvidia are working on now, so logically it makes sense that Apple’s production processes are part of the plan.
Designed by Apple, built by robots
Achieving this is not going to be easy. But where Apple goes, others inevitably follow, which itself means that future employment is going to become even further deindustrialized at about the same time as AI itself leads to mass scale changes in working practices elsewhere. It’s hard to see where this is going, but the other side of that story is that iPhone manufacturing will itself become a movable feast.
“Designed by Apple, built by robots,” some might say.
You can follow me on social media! Join me on BlueSky, LinkedIn, Mastodon, and MeWe.