South Carolina has assembled an unusual concentration of advanced manufacturing in a state its size — and a credible claim to becoming the Southeast's applied industrial-AI capital. The BMW–Figure AI humanoid-robotics pilot is the most visible signal of what is actually underway. The deeper question is how the state's manufacturers, universities, and workforce system convert a strong assembly of capabilities into a recognized national position.
The most-watched manufacturing-AI development in South Carolina in the past year is also the easiest to underestimate. In early 2025, BMW announced that the Spartanburg plant would host a pilot deployment of Figure AI's humanoid robots — making BMW Spartanburg one of the first commercial sites in the world to install bipedal humanoids on a real production line. The pilot is small. The headlines were larger. The signal — that BMW chose its largest plant globally to test a new class of industrial automation — is what matters.
The Spartanburg pilot is one window into a broader story. South Carolina is not a state that talks about itself as a manufacturing-AI hub, but in the assets it has assembled, it has more credible material to make that claim than most other states. This article walks through what those assets are, what is actually happening on the plant floor today, and what would help South Carolina's manufacturers turn applied capability into a recognizable national position.
Manufacturing is South Carolina's signature sector. The state ranks among the top in the country for manufacturing share of state GDP, foreign direct investment in manufacturing, and per-capita exports. The visible names anchor the picture, but the depth comes from the tier-one and tier-two supplier ecosystem that has grown up around them.
The Spartanburg plant produces X-series SUVs for global export and employs more than 11,000 South Carolinians directly, with multiples of that number in the supplier base. In 2025, BMW began piloting Figure AI's humanoid robotics on a real production line — among the earliest commercial deployments of bipedal humanoids anywhere.
Boeing's 787 final-assembly operation is the only Boeing widebody assembly line outside the Pacific Northwest. The site employs roughly 6,500 workers and is among the more digitally instrumented aircraft-assembly operations in the world.
Volvo's $1.1 billion Ridgeville plant is the company's only US manufacturing site, producing electric vehicles for the North American market. It has been a public-facing reference site for Volvo's Industry 4.0 initiatives.
The Sprinter assembly plant produces the workhorse of America's commercial-van fleet. Its computer-vision quality-control systems are among the more visible examples of inline machine-learning applications in the state.
Michelin's North American headquarters and several manufacturing operations make the company one of the state's largest private employers. Predictive-maintenance and computer-vision quality-control work has been a recurring theme in Michelin's public communications about the SC operations.
The tier-one and tier-two supplier ecosystem — anchored by tire and chemical producers, suppliers to BMW and Volvo, and a growing number of EV battery and component producers — is where much of the day-to-day applied-AI deployment in SC manufacturing happens.
The conversation about AI in manufacturing tends to fixate on the dramatic — humanoid robots, "lights-out" factories, autonomous mobile robots. The day-to-day picture is less dramatic and, in productivity terms, more interesting. Most of the applied AI in SC manufacturing today is concentrated in four areas, each of which is generating measurable returns somewhere in the state.
Sensor-rich production equipment is producing data that, fed into machine-learning models, can predict failures before they happen. Michelin has publicly discussed predictive-maintenance work; tire production is unusually well suited to it because individual machines are large, expensive, and continuously instrumented. Predictive maintenance is the most consistent productivity story across SC's heavy-manufacturing operations and is one of the clearer examples of AI's effect on the bottom line being routine rather than novel.
Inline computer vision for defect detection is now common at SC's automotive and aerospace operations. Mercedes-Benz Vans, Volvo, BMW, and Boeing have all publicly described computer-vision applications in body, paint, and assembly quality control. The pattern is consistent: traditional manual or fixed-fixture inspection is being supplemented (and in some cases replaced) by camera-and-model systems that catch defects earlier and more consistently. The result is fewer rework cycles and tighter feedback loops to upstream processes.
Material handling, parts movement, and increasingly assembly itself are being supplemented by autonomous mobile robots and traditional industrial robots equipped with newer perception systems. The BMW–Figure AI pilot is the most visible example, but it is not the only one — most of SC's larger plants are running AMR pilots of some scale. The economic case is robust where labor markets are tight, which describes much of the SC manufacturing corridor.
Behind the scenes, machine learning is reshaping how SC manufacturers think about logistics, inventory, and supplier coordination. The Port of Charleston's continued digitization (covered in a related Journal piece on infrastructure) is one piece. Inbound-logistics optimization at BMW, Volvo, and Boeing is another. The Inland Port at Greer makes the Upstate's freight infrastructure unusual: AI-driven coordination between port operations and inland-port operations is one of the places SC's logistics edge could compound.
"The thing the rest of the country misunderstands about South Carolina manufacturing is the depth. The headline names are real, but the supplier ecosystem around them is what makes this state different. That ecosystem is where most of the applied AI is happening — quietly, and faster than anyone outside the state realizes."
BMW's pilot deployment of Figure AI's humanoid robots at Spartanburg is the development that has drawn the most national attention. It is worth understanding what the pilot actually is and what it is not.
Figure AI is a Sunnyvale-based humanoid-robotics company building general-purpose bipedal robots — the kind of system science fiction has trained us to imagine, now real enough to deploy in production environments. The BMW pilot is small in scale: Figure's robots are being trained and evaluated on specific, well-defined tasks at the Spartanburg plant, not replacing assembly workers. The productivity case at this stage is not the point. The strategic significance is the choice to do this at Spartanburg.
BMW Group runs more than 30 manufacturing plants worldwide. The decision to pilot the most experimental class of factory automation at the company's largest plant by volume — and to do it in South Carolina rather than Munich, Leipzig, or Dingolfing — is a signal about how BMW thinks about the Spartanburg operation. It is also a signal about how Figure AI thinks about the kind of plant where its technology can be developed productively. That kind of multi-year industrial relationship between an OEM and a frontier-AI hardware company is a thing that other states will spend the next several years trying to recruit.
Applied industrial AI does not happen in a vacuum. It depends on a research substrate that turns out engineers, develops new techniques, and provides the credibility that makes industry-university research collaboration possible. South Carolina's substrate here is stronger than the manufacturing base alone suggests.
Clemson's International Center for Automotive Research (CU-ICAR) in Greenville is a flagship industry-academic facility — Clemson's automotive engineering graduate program is on-site, and the campus has long-standing relationships with BMW, Michelin, and the broader Upstate automotive supplier base. CU-ICAR's research portfolio covers autonomous systems, vehicle dynamics, advanced manufacturing, and increasingly applied AI in industrial contexts.
USC's McNair Aerospace Center serves the Boeing relationship and the broader Lowcountry aerospace ecosystem. The South Carolina Manufacturing Extension Partnership (SCMEP), part of the NIST-funded national MEP network, provides hands-on technical assistance to smaller manufacturers — and is one of the natural channels for AI adoption to reach the tier-two and tier-three supplier base that the OEMs depend on. The SC Department of Commerce's apprenticeship and workforce programs round out the picture.
The pieces — large OEMs, deep supplier base, two research universities with industry programs, a state-level workforce machine, and an established applied-research institution in CU-ICAR — are unusual in their density for a state of South Carolina's size.
The most common question about manufacturing AI is whether it will displace workers. The honest answer for South Carolina is: in net terms, probably not on the timeline that gets discussed most, but the composition of manufacturing employment will shift faster than most people expect. The harder question is whether the state's workforce-training infrastructure can keep up with the shift.
Applied AI in manufacturing creates new operational roles even as it shifts the contents of existing ones. Data labeling and quality-assurance roles, model-output review, MLOps support for the manufacturing IT stack, applied prompting for shop-floor decision support, and AI-aware project management are all jobs that exist on SC plant floors today but did not exist five years ago. None of them require four-year degrees. All of them require structured short-form training that the state's Technical College System is well-positioned to provide.
ReadySC, the SC Technical College System's employer-customized training program, has been one of the most consistent advantages in the state's manufacturing recruitment for decades. The opportunity for the next decade is to make ReadySC and the broader Technical College System the primary engine for AI-adjacent manufacturing credentials in the Southeast. The infrastructure exists; the specific credentials and curricula need to be developed and stacked into recognizable pathways.
South Carolina does not need to invent industrial AI; the pieces are in place. The constructive question is how to convene and connect them.
For readers tracking how SC's manufacturing-AI story develops, five specific signals will tell most of the story over the next two years.
SCAIO's posture, on this question and others, is to describe what is and what would help, not what should happen. The work to make South Carolina the Southeast's applied industrial-AI capital is, in nearly every case, already underway in pieces somewhere in the state. The work that remains is the work of connection — between the universities and the manufacturers, between the OEMs and the supplier base, between the Technical College System and the deployment sites, between the in-state story and the national audience.
Manufacturing has been the cornerstone of South Carolina's economic identity for a generation. AI in manufacturing is on track to be one of the cleanest productivity stories of the next decade. The state is positioned well to lead, and the institutions doing the work deserve more national visibility than they currently have. SCAIO will continue to document the story as it develops.