This blog follows on from our recent post on radical disruptions to hardware engineering
By Sam Smith-Eppsteiner and Thilo Braun
Manufacturing, a centuries-old industry historically slow to embrace change, is now being rapidly propelled into the 21st century. Supply chain disruptions and geopolitical tensions have exposed the vulnerabilities of outsourcing, while labor shortages across the US and Europe further strain an already pressured sector. Meanwhile, breakthroughs in AI and robotics signal an impending wave of automation. The industry is clearly on the cusp of a profound transformation—but what will manufacturing look like once the dust settles?
Note: manufacturing is a behemoth sector - ranging from chemicals and food to semiconductors and fighter jets. This post will focus on manufacturing discrete products and systems such as aircraft, robots, rockets, cars and their components.
The sector has long been a cornerstone of geopolitical strength. During the Gilded Age and throughout the world wars of the 20th century, industrial capacity was directly tied to national security. The ability to mass-produce tanks, planes, and weapons reshaped world maps and determined military dominance.
The US’s ability to repurpose large swathes of civilian manufacturing to produce military equipment, such as Ford and General Motors assembly lines producing fighter-planes and tanks, tipped the balance of power in World War II. Conversely, both Russia and Ukraine are heavily dependent on China’s production of components for drones to power their respective armies in the ongoing war in Ukraine.
While manufacturing may not dominate today’s headlines, it remains central to geopolitical power. It underpins the development of critical technologies—from AI and advanced computing to energy infrastructure—that societies increasingly rely on.
However, the West, and in particular the US, have increasingly let manufacturing move to lower cost countries such as China. In 1980, the United States manufactured over 40 percent of global high-technology goods, compared to just 18 percent today (Atkinson). US manufacturing now makes up only 10% of GDP vs 20% in Germany and 26% in China (NIST, DESTATIS, China Briefing).
Both sides of the aisle in the US agree that it's time to re-industrialize. The goal is not to restore low-skilled manufacturing lost to globalization in the late 20th century, but to lead in AI, robotics, and advanced production technologies—securing economic and strategic dominance in the 21st century.
Real Total Manufacturing Construction Spending (Source: US Treasury)
With the passing of the CHIPS Act and the Inflation Reduction Act, we’ve already seen a significant uptick in manufacturing investment in the US. However, we believe this is only the beginning, and new technologies are critical to support this manufacturing renaissance in the US.
Our conversations with manufacturing executives have surfaced several themes :
The challenges outlined above create opportunities for new innovation. We see opportunities across new manufacturing technologies, supporting manufacturing operations, and even in building full-stack manufacturing capacity.
The US has significant manufacturing capacity that is largely underutilized for lack of labor. At the same time, companies are shifting manufacturing to other countries because they are failing to secure sufficient labor. To alleviate the industrial workforce challenge, new approaches are needed.
We’re looking for founders with innovative approaches to help resolve the acute labor shortage we are facing today. This is the number one problem for many industrial leaders and the time is ripe for new technologies to be introduced. We see significant opportunities in leveraging AI for RAG / knowledge use cases as well as agentic workflows
The convergence of AI and robotics is creating a new realm of opportunities. Programming robots has been a key constraint to deploying automation at scale. Automation has been most useful in low-mix, high volume opportunities.
Robotics serves high volume, low mix applications today. We anticipate increasing use in high volume, high mix applications before becoming omnipresent across all volumes and product mixes.
Artificial intelligence is enabling increasing flexibility in automation. This will initially translate to serving a higher mix at high volumes. Over time, we anticipate automation becoming omnipresent even in low volume applications.
Advanced AI models are enabling robots to handle variable tasks, adapt to changing production lines and products, and work safely alongside humans (e.g., Physical Intelligence, Covariant).
Applying this enables lower cost automation (e.g., RobCo, Launchpad) and new robotic form factors (e.g., Figure, 1X Technologies).
We’ve written extensively about our experiences with robotics (eg. Lessons learned the hard way: Robotics and The art of automation: original mistakes with robotics innovator Jorge Heraud) and this topic deserves a blog post of its own. We’ll keep it short here: we’re excited by a new generation of robotics emerging to power the future of manufacturing.
New technologies are enabling novel manufacturing processes. Additive manufacturing has become commonplace in some key applications including jet engine components, casting moulds, and spare parts manufacturing. In addition, we are seeing novel automation opportunities such as Machina Labs reinventing sheet metal manufacturing and enabling new product capabilities. Atomic Industries is leveraging AI to support tool and die design and manufacturing. Quantum Diamonds is creating new ways of sensing otherwise undetectable defects in manufacturing semiconductors.
We’re excited by ambitious founders utilizing recent technology advancements to create new ways of making things, often enabling product capabilities that would otherwise not be manufacturable.
Artificial intelligence has taken the world by storm with many potential applications in manufacturing. The Special Competitive Studies Project put together a great action plan for advanced manufacturing with a focus on AI. Labor, robotics, and novel processes described above rely on AI in many cases, and there are many more applications not captured here.
Generative AI is already used in back-office tasks such as managing supply chains and quoting, led by innovative companies like Toolpath, Didero, and Soff. Increasingly, we expect this to translate to the shop-floor, augmenting machine programming, troubleshooting, and error prevention.
The previous generation of machine learning for manufacturing has taught us that achieving adoption can be challenging and that pilot purgatory is real. Founders need deep empathy for the customer and initially tackling enabling/peripheral processes accelerates adoption. It’s easier to sell a brewery AI for predictive maintenance and troubleshooting than anything that might affect the taste of the beer.
In addition to technological advances, we are looking at companies thinking creatively about business models. Deploying new manufacturing technologies in existing industrial supply chains can be challenging for various reasons including low risk tolerance, high sunk cost in equipment, constrained capital amongst component manufacturers, and the need for new engineering approaches to utilize novel manufacturing technologies.
In order to successfully deploy novel technologies or processes, vertical integration may be required. Companies such as Re:Build Manufacturing, Daedalus, and Hadrian are spearheading new manufacturing capacity build-out.
We are looking for companies with a strong customer understanding and commercial DNA.
Manufacturing in the West is experiencing a renaissance. To be globally competitive, we need to manufacture faster, more flexibly, cheaper, and greener. AI and robotics hold the key.
We’ve invested in leading companies innovating in manufacturing including Machina Labs, Form Labs, Fero Labs, and Citrine Informatics. We’re excited to talk to builders building for makers — we see a huge opportunity in enabling and accelerating the future of manufacturing.
