ACT Expo 2025: Battery Experts Fret Tariffs, Encourage Specific Designs

A panel of battery makers grappled with the right response to political and economic changes, called for more specialized cell chemistries and pack designs for unique use cases and embraced AI for predictive failure analysis.

April 30, 2025

Chris Clonts

There’s a chaotic elephant in almost every room of the 2025 Advanced Clean Transportation (ACT) Expo: Tariffs and the Trump administration’s seemingly hostile approach to environmentally friendly technology like EVs and alternative energy sources.

Panelists for the “Cells to Solutions” session, from left: Selman Tosunoğlu, founder, CEO and CTO of Ampherr AG; Paul Beach, president of Octillion Power Systems; Kevin Dahlberg, VP of cell technology for Freudenberg e-Power Systems; and Trevor Grant, cell development manager for Infinity battery maker Electrovaya. (Chris Clonts)

The Trump administration’s announcement on the second day of the ACT Expo that it would be lowering auto tariffs only underscored the uncertainty.

During a panel session titled “Cells to Solutions” at California’s Anaheim Convention Center, Paul Beach, president of Octillion Power Systems, said that despite the stated goal of tariffs to move factory jobs back to America, the whipsaw instability of policy is making most executives freeze rather than make the leap to build factories in America.

Beach said that because 70% of Octillion’s revenue is in China and 30% is in India, the company isn’t being affected much by the tariff uncertainty. “I don’t think any one of our customers is willing to go fight the administration right now in a lawsuit. So we have to face that problem,” he said.

“There are cells produced in the U.S. So we can source cells here. We have cell vendors we’re working with. I can get cells from Taiwan. I can get cells from Korea. I can send them to the U.S., and I can package them. But,” Beach said, “the reality is it’s going to be two or three times what the cost is when you build the packs in China. So even at the 200% tariff, it’s just the reality of building things in the U.S. It’s more expensive.”

Selman Tosunoğlu, founder, CEO and CTO of Ampherr AG, said that the company builds its batteries in the United States, Germany and Turkey, so tariffs are a minor hurdle. “We believe that we should definitely start with production in the United States, but that is just kind of solving partially that problem,” he said. “Another thing is also diversifying the source of the cells,” he said, adding that they are working to be flexible and open to switching suppliers.

But even switching suppliers, in the current climate, is a risk due to the unstable nature of tariffs and regulations, other panelists agreed.

Trevor Grant, cell development manager for Infinity battery maker Electrovaya, said it would be exceedingly difficult to produce cells in the United States unless they come with some sort of competitive advantage. “The customer needs a reason to be buying the cell. If you’re trying to produce the same cell domestically, there’s no reason because it’s the same cell, just 50 percent more expensive,” he said. “So our focus really is on the cell technology side. The customer’s reason to buy is either for tariff reasons, greenhouse reasons or technology reasons.”

Embracing task-specific designs

As for what battery solutions would look like in the future, Electrovaya’s Grant suggested that the industry needs to be more willing to explore even more exclusive designs and chemistries to match applications, especially in the off-highway and industrial spaces. He described the life of a typical passenger EV battery as easy, typically using less than 15 percent of its capacity before being recharged. That translates to a battery built to last 1,000 cycles, or roughly 20 years – the life of the vehicle in most cases.

“If you think about warehousing and the forklift. This is in the distribution center, the warehouse of a Fortune 100 company. These are basically 24/7 operations. That battery is being used continuously, 365 days a year. Sometimes doing up to 1.5 or two cycles per day, and at that rate, we’re at 1,000 cycles in under three years, right? This is a huge difference in usage,” Grant said. “Can we expect the same battery, the same cell technology in the space to be used in both of those applications? And can we expect a battery that’s designed for passenger EVs to perform well in these demanding applications? My answer to that question is no, and it’s very important, I think, to be matching our cell technology to the application.”

Paul Beach, president of Octillion Power Systems, was emphatic about battery makers tailoring cells and chemistries to specific applications. (Chris Clonts)

Octillion’s Beach agreed and said potential customers often approach him and cite Tesla as the model they want to follow for electrification. “Whether it’s a truck guy, car guy, Caterpillar, they all use Tesla as the example. Tesla packs: You can fast-charge them, you can use 95% of the capacity, it’s going to have an eight-year warranty and it’s $125 per kilowatt hour. Give me that,” he said. “And I’m like, let’s start from a different conversation. What is it you build? Do you build a Tesla car? Are you building a passenger car? No, you build a truck. How’s that truck used? That truck is used every day and driven to 100% of its capacity.”

Beach continued, saying that customers want to fast-charge their trucks. “But people don’t fast-charge Teslas every day. If they do, they will destroy that battery.” He called what the potential customers really want is “unobtanium.” He said customers should keep an open mind and target the minimum viable product for each application.

The role of AI

The panel enthusiastically endorsed AI for use in predictive failure analysis and diagnostics down to the cell level, such as that pioneered by Voltaiq. But Kevin Dahlberg, VP of cell technology for Freudenberg e-Power Systems, cautioned that setting up that kind of system required deliberate steps.

“AI isn’t just something to turn on and then you’re using it. Before you get to data science there needs to be a very robust data engineering initiative,” Dahlberg said. “You need to be able to structure validation from the field, from the lab, from manufacturing, all into one place. Is your BMS [battery management system] and your telemetry capable of getting the data to the database? And having good data structure itself is critical.”

Arun Chickmenahalli, director of maintenance R&D at Ryder System, moderated the discussion.