The most critical shortage in American heavy manufacturing isn't a material or a chip. It's a person. For years, projections have warned of a deficit of hundreds of thousands of skilled welders, a gap that constrains production from shipyards to defense contractors and introduces costly quality variability [The Robot Report]. In Columbus, Ohio, a company founded by two brothers is betting the answer isn't just a better training program, but a new class of worker altogether: one made of steel, silicon, and artificial intelligence.
Path Robotics, founded in 2018 by Andy and Alex Lonsberry, calls its approach "Physical AI." It's a label that encompasses the company's full-stack integration of computer vision, proprietary AI models, robotic hardware, and autonomous motion control [Path Robotics]. The initial, and still primary, application is autonomous robotic welding, a task the company deliberately selected as one of the hardest factory-floor jobs to automate. By starting there, Path aimed to stress-test its systems against extreme industrial complexity, from variable part fit-up to changing thermal conditions [BriefGlance.com]. The bet appears to be paying off in both contracts and valuation. In November 2026, the company announced a $300 million Series E round led by Georgian, catapulting its post-money valuation to $2.45 billion and bringing its total disclosed funding to approximately $610 million [Path Robotics, November 2026][Bloomberg, November 2026].
The Wedge of Welding
Path's technology centers on its Obsidian AI, a foundational model the company says was trained on "tens of millions of welded inches" within a simulated environment it calls the Weld World Model [Path Robotics][SiliconANGLE, November 2026]. Unlike traditional robotic welding that requires precise programming and fixturing for each part, Obsidian-equipped systems use cameras and lasers to scan each unique weld joint in real time. The AI then generates a custom execution plan, adapting to the part as it actually exists, not as it was designed [techdoaction.com]. The company claims this allows its AF-1 and AW-3 robotic cells to run autonomously for hours, achieving a near 99% first-pass yield compared to a human welder's typical 60-70% [Machine Design][ManufacturingTomorrow].
This performance is packaged for customers through two key models. The first is a hardware-as-a-service (HaaS) or robotics-as-a-service (RaaS) subscription, where Path retains ownership of the equipment and customers pay a recurring fee, often tied to productivity and quality guarantees [Industrial Monitor Direct]. The second is a dedicated "Mission Control" support team, providing 24/7 remote assistance to maximize cell uptime [Path Robotics]. This combination is designed to lower the barrier to adoption for manufacturers wary of large upfront capital expenditures and complex integration.
Traction and the HII Lighthouse
Customer adoption provides the clearest signal that Path's systems are moving beyond pilot projects. The company's reported annual recurring revenue grew 17-fold in the 18 months leading up to its Series E announcement [Securities.io]. Its workforce surged 97% year-over-year, with plans to reach 400 employees by the end of 2026 [Business Wire, November 2026].
The most significant validation, however, is a performance-based production agreement with HII, one of America's largest military shipbuilders. Announced in 2026, the deal is worth up to $600 million and aims to accelerate the deployment of Path's systems across U.S. Navy shipbuilding programs [Robotics 24/7, November 2026]. This partnership is a lighthouse customer in the truest sense, proving the technology in one of the most demanding welding environments imaginable. "Welding is one of the most complex processes to automate in any industry, and shipbuilding raises that bar even higher," a company statement noted [Markets Insider, January 2026]. Other named customers include Lockheed Martin, Saronic, and several mid-sized industrial manufacturers [Robotics 24/7].
2024 Series D | 100 | M USD
2026 Series E | 300 | M USD
The Expansion Beyond the Weld
With welding as its proven beachhead, Path's stated strategy is to expand into adjacent factory tasks. The company plans to move into processes before and after welding, such as assembly, inspection, grinding, and painting, leveraging the same Physical AI stack [Path Robotics, November 2026]. A glimpse of this future is already visible in Rove, a mobile welding robot that pairs the Obsidian AI with a quadruped platform. Rove is designed to bring automation to large, unstructured environments like shipyards without the need for fixed cells or extensive fixturing [Path Robotics].
The company's ambition is framed as a broader shift in where AI's center of gravity will reside. "The AI center of gravity will ultimately shift from screens to the physical world, including manufacturers, shipyards, job sites and beyond," wrote a Path executive in a Forbes Council post [Forbes, March 2026]. In this view, AI that can weld, assemble, and build demonstrates tangible economic impact where it matters most: on the factory floor.
Risks in a Capital-Intensive Race
The scale of Path's ambition is matched by the scale of its challenges. Building and deploying physical robots is profoundly capital-intensive, as evidenced by the $610 million already raised. The company must now execute on its expansion roadmap while managing the complexities of scaling hardware production, global service logistics, and a rapidly growing team.
- Execution complexity. Moving from welding into entirely new manufacturing processes like assembly or painting requires significant new AI training data and potentially new hardware form factors. Each new task represents a fresh engineering mountain to climb.
- Economic sensitivity. Path's subscription model ties its revenue to customer production volume. In an economic downturn where manufacturing output slows, recurring revenue could face headwinds, testing the resilience of the business model.
- Competitive response. The company is not alone in pursuing factory automation. Established robotics incumbents and specialized welding automation firms like Novarc and ARC Specialties are formidable competitors. Path's differentiation rests on its full-stack AI integration, but that lead must be continuously maintained.
For now, the company's war chest and flagship customer provide a formidable runway. The recent funding is earmarked for scaling operations, accelerating R&D, and meeting the demand signaled by deals like the one with HII [Path Robotics, November 2026].
The Standard of Care on the Factory Floor
The patient population here is not defined by a disease code, but by an industry under strain. It is the U.S. manufacturing base, and specifically the heavy industries,shipbuilding, defense, structural steel,that have relied for decades on a dwindling cohort of master welders. The standard of care today is a painful trade-off. Manufacturers can struggle with the high cost, long lead times, and quality inconsistency of a scarce human workforce, or they can invest in traditional robotic automation that is often inflexible, expensive to program, and ill-suited for high-mix, low-volume or large-scale work.
Path Robotics is proposing a third path: autonomous systems that see, think, and adapt in real time. The next twelve months will be about proving that the precision demonstrated in welding cells can be replicated across the broader factory floor, turning a compelling bet on Physical AI into an industrial reality. The company's trajectory suggests it is no longer just selling a robot, but a new way to build.
Sources
- [Path Robotics, November 2026] Path Robotics Announces $300 Million Series E Financing at $2.45 Billion Valuation | https://www.path-robotics.com/news/path-robotics-announces-300-million-series-e-led-by-georgian
- [Bloomberg, November 2026] Path Robotics Quadruples Valuation to $2.5 Billion in Latest Funding Round | https://www.bloomberg.com/news/articles/2026-11-10/path-robotics-quadruples-valuation-to-2-5-billion-in-latest-funding-round
- [Business Wire, November 2026] Path Robotics Announces $300 Million Series E Financing | https://www.businesswire.com/news/home/20261110046181/en/Path-Robotics-Announces-$300-Million-Series-E-Financing-at-$2.45-Billion-Valuation-to-Scale-Physical-AI-for-Manufacturing
- [Robotics 24/7, November 2026] Path Robotics announces $300M Series E financing at $2.45B valuation | https://www.robotics247.com/article/path_robotics_announces_300m_series_e_financing_at_2.45b_valuation
- [The Robot Report] The demand for welders is increasing at 3% per year | https://www.therobotreport.com/
- [Securities.io] Path Robotics reported 17-fold ARR growth | https://www.securities.io/
- [Markets Insider, January 2026] HII Launches HYPR Program with Path Robotics | https://markets.businessinsider.com/news/stocks/hii-launches-hypr-program-with-path-robotics-and-graymatter-robotics-to-accelerate-production-at-scale-1036039456
- [Machine Design] Path Robotics welding solutions achieve a near 99% first-pass yield | https://www.machinedesign.com/
- [Forbes, March 2026] Council Post: The Future Of AI: Top CES Takeaways | https://www.forbes.com/councils/forbestechcouncil/2026/03/11/the-future-of-ai-top-ces-takeaways-from-tech-experts/
- [BriefGlance.com] Path stress-tested its AI against extreme industrial complexity | https://briefglance.com/
- [techdoaction.com, 2026] Obsidian AI scans each weld joint as it actually exists | https://techdoaction.com/
- [SiliconANGLE, November 2026] The Weld World Model generates simulated welding projects | https://siliconangle.com/
- [Industrial Monitor Direct] Path's automation-as-a-service model with fixed monthly fees | https://industrialmonitordirect.com/
- [ManufacturingTomorrow] AF-1 cells can run autonomously for hours | https://www.manufacturingtomorrow.com/