Finding the Missing Piece of America’s Drone Supply Chain

Inside SES.AI, advanced battery chemistry, AI-driven materials discovery and global manufacturing illustrate the challenge of building the industrial base beneath America’s growing drone market.

On the factory floor at SES.AI’s Massachusetts facility, the challenge of building a secure U.S. drone supply chain becomes tangible. Batteries are being developed for applications where weight, performance and reliability matter enormously. But next door, in the office of founder and CEO Dr. Qichao Hu, a much smaller object helps explain where battery technology may be going next.

Hu’s office reflects the mind of a scientist: shelves of advanced technical books, along with a complete set of Douglas Adams novels. Nearby sits an unassuming metal box, roughly six inches square and two inches deep.

The box holds access to what SES.AI calls its “Molecular Universe,” a database mapping roughly 100 million molecules that could offer useful properties for battery development.

The contrast is striking. A small box represents an enormous range of possible materials.

And in an unassuming building in a Boston suburb, SES.AI itself illustrates an equally large challenge facing the U.S. drone industry: building and scaling the battery and critical component ecosystem needed to support growing domestic drone production.

The Ecosystem Behind the Aircraft

U.S. drone policy increasingly emphasizes secure supply chains and reduced dependence on Chinese components. But replacing a foreign-made drone with a U.S.-made aircraft addresses only part of the problem.

Every drone depends on an underlying ecosystem of batteries, motors, electronics, sensors and other components. Rebuilding that ecosystem is more complicated than opening an assembly plant.

Hu provides a clear view of the issue.  “If you look at the evolution of the battery industry, we lost a lot of the capabilities in the US decades ago,” Hu said. First the expertise and manufacturing moved to Japan, then to South Korea and China.

Now that US drone manufacturers cannot rely on sourcing from China, they have turned to other allies with the right set of capabilities. In Chungju, South Korea, where SES.AI operates a manufacturing facility, Hu describes a cluster of manufacturers supported by “an ecosystem of talent and suppliers.” SES AI identifies Chungju as its newest manufacturing facility and says it complements the company’s other global operations.

The distinction matters. A factory can be built. Recreating a network of experienced workers, materials suppliers, equipment makers and specialized knowledge takes much longer.

For the U.S. drone industry, however, the need for alternatives is becoming increasingly urgent.

From Drones to EVs, and Back Again

Drones are not a new market for SES.AI. Hu said the company focused on drones and other aviation applications as early as 2018 and 2019.

“Our batteries are high energy density and super lightweight, so anything that flies is a great application,” he said.

Then the market changed. During the COVID period, investment in electric vehicles accelerated. Large SUVs and pickup trucks presented another problem where battery energy density mattered: range. SES.AI attracted significant investment from automotive manufacturers and shifted much of its attention toward EVs.

Now, however, drones are again one of the company’s three focus markets as they grow and expand.

That detour through automotive manufacturing has proven valuable. Hu said automotive customers impose substantially more rigorous manufacturing requirements than drone customers. He estimates more than 3,000 quality checkpoints for automotive cells, compared with about 400 for drone applications.

SES.AI is therefore returning to aviation with manufacturing processes shaped by years of work with the automotive industry.

The company is also no stranger to secure sourcing.  Well before today’s drone supply-chain push, SEI was working with GM Defense – which drove the development of manufacturing capacity in South Korea.

Policy Creates New Urgency

The changing U.S. policy environment has altered the calculation for drone manufacturers, taking some by surprise.

Hu said many companies delayed major sourcing changes while federal policy remained uncertain.

“Most drone customers just stock-piled Chinese made cells in 2024 and 2025.  They took a ‘wait and see’ approach because the policy was not clear,” he said.

That situation has changed as federal requirements around Chinese drones and critical components have become clearer.  “Now, the issue has become much more urgent,” Hu said. “It has been very helpful to have clear guidelines.”

For the developing U.S. drone industrial base, restrictions on Chinese products can create demand for alternatives.  But policy alone cannot create the manufacturing capacity required to supply them. Companies must still develop compliant components and produce them at commercial scale.

Why Batteries Matter for Drones

For electric vehicles, better batteries can mean greater driving range. For drones, the weight equation is even more unforgiving. Every gram devoted to the battery is weight that cannot be used for a camera, communications system, medical shipment or other payload.

Hu describes the potential impact of SEI’s higher energy density batteries simply: “An aircraft of the same weight can fly twice as far – or fly the same distance with double the payload.”

Actual performance depends on the aircraft and mission, but the underlying principle is straightforward. Improving the amount of energy available at a given battery weight can change what an aircraft can carry and how long it can remain airborne.

Finding the materials that can deliver those improvements, however, presents another enormous challenge.  SES.AI’s approach to the problem was bold – and successful.

Exploring a Molecular Universe

Battery development has traditionally been slow.  Battery electrolytes use small organic molecules: different applications or outcomes require different characteristics. One customer may need a battery that performs in extreme cold. Another may prioritize heat tolerance or energy density.  Choosing the right material is not simple.

The number of possible molecules is vast. After eliminating unstable and unsuitable possibilities, Hu said SES.AI was still looking at a universe of approximately 10^8, or 100 million, molecules. By comparison, he estimates that the battery industry historically explored only around 1,000.

Testing those possibilities through conventional research would be impractical.

That’s where SES.AI turned to artificial intelligence.

“We didn’t do AI for the sake of AI – it was for product development,” Hu said.

Working with NVIDIA, SES.AI used AI and computational tools to map the properties of its Molecular Universe. Hu said the company completed that mapping in 2025.

“Before AI, that could have taken thousands of years – but now, you can map all of those properties in about a year,” he said.

The goal isn’t simply to accumulate an enormous database. SES.AI uses that information to narrow the search for materials suited to a particular application.

“Most material discovery takes a long time, and your chance of success is very low. AI accelerates that,” Hu said. “Instead of taking years, it now takes weeks.”

The system can evaluate possibilities, learn from results and recommend promising candidates. That could allow battery developers to respond much more quickly when a customer or an application needs different performance characteristics.

The small metal box in Hu’s office makes the concept unexpectedly concrete. What resembles an ordinary piece of computer hardware provides access to a digital encyclopedia of potential battery materials – now available commercially.

The Next Problem Is Scale

Finding better materials solves only part of the problem. They still have to be manufactured.

SES.AI currently has capacity in South Korea to produce about one million cells annually. Hu says that isn’t nearly enough to address the demand the company sees developing. SES AI has publicly described plans to expand its Chungju capacity to approximately one million Li-Metal and Li-ion pouch cells annually to meet demand from U.S. and European drone customers.

“The potential demand is almost 100X that,” Hu said.

SES.AI is now looking at additional capacity in Japan and Malaysia.

That gap between available production and potential demand captures the larger challenge facing the U.S. drone industry.

American companies can design aircraft, develop autonomy and assemble drones domestically. But a rapidly expanding drone industrial base also requires batteries and other critical components in quantities that today’s secure supply chains may not yet be able to provide.

The most visible evidence of America’s drone expansion will be new aircraft and new factories. Much of the harder work, however, is happening deeper in the supply chain.

SES.AI offers a glimpse of that less visible industrial base: advanced materials research in Massachusetts, manufacturing expertise developed through the automotive industry, and production capacity spread across trusted international supply chains.

For U.S. drone manufacturers, developing that component ecosystem may be just as important as building the aircraft themselves.

Read more:

• The Drone Industry’s Next Bottleneck Isn’t the Aircraft
• U.S. Drone Manufacturing’s New Advantage Isn’t Just Factories. It’s Agility.
• ePropelled Secures $60 Million in U.S. Government Funding to Expand Drone Propulsion Manufacturing