Automated Testing Key to Battery Development Panelists: Grid Upgrades Needed Before Full EV Adoption

TDK Ventures
TDK Ventures
Published in
6 min readJan 22, 2022


While electric vehicle development and adoption continue to advance, EV charging infrastructure in the U.S. remains in its infancy. Not only will more charging stations soon be required but academia and industry also need to develop faster-charging technologies. We must take steps to upgrade the grid, implement microgrid buffers, and expand our electrical infrastructure to accommodate these new demands. Day 3 of Energy Week presented by TDK Ventures assembled by some of the best and brightest individuals from all aspects of EV and battery research to discuss the state of the industry, pending breakthroughs, and the challenges surrounding electromobility and grid issues.

Rajesh Swaminathan, Partner at Khosla Ventures, kicked off the discussion by calling for a more structured, machine language (ML) and artificial intelligence (AI)-driven approach to designing and evaluating battery performance, materials, and chemistries. He said in his 13 years of investing in battery technology, most have used a “brute force” method.

“They bring really smart people together to figure out what electrolyte chemistry works best, how to optimize for applications. It is time to leverage so much that has happened in machine learning, AI, and even in the next decade, quantum computing to figure out — and in some ways start from scratch — to see what the ideal combination of material choices is we have to be making and what kind of performance parameters we should be shooting for.”

Ju Li, material science and engineering professor at MIT, noted that his team is trying to tackle this specific problem, working with complex electrolytes with multiple solvents and high-concentration salts to improve the cycle life and safety.

“Some of the difficulty comes because we don’t have a very good way of testing these formulations and how they will behave in the cell setting. There is a lot of manual labor involved. If we can automate cell-making or to test without even making a coin cell, that would be great.”

Approaching battery development and applications using AI/ML could generate advances like those driven by similar approaches in the pharmaceutical industry, noted Veronica Li, Partner at Primavera Capital Group. She said those drug discovery insights have been realized thanks to automation and technologies that can uniformly categorize, label, and tag vast amounts of data across several databases.

Julia Song, co-founder and chief technology officer at Energy Storage Systems (ESS Inc.), cautioned against focusing solely on performance, especially when scaling from lab proofs to implementation and “manufacturing something that is cost-effective. Our work with national labs has shown promise in getting to better performance, durability, and cyclability, but there’s no data behind how much manufacturing cost you have to put in place to support these incremental efficiencies and whether it justifies the work you are trying to do.”

Finding the answers may rest on a company’s ability to attract diverse contributors from research universities and a range of verticals working to commercialize battery technologies, said Sarah Applebaum, Partner at Pangaea Ventures.

“The team that you need to run your AI/ML discovery platform and optimization is very different from the team skillset and resources you need to build, fund, operate, and execute on a battery materials company,” she noted. “How do you marry the discovery engine with the execution go-to-market?

As the conversation then turned to electricity infrastructure, most panelists expressed pessimism that extensive adoption of vehicle-to-grid applications for moderating demand loads and providing backup power will come to fruition in the near future.

“It may occur in a pretty limited extent in certain niche scenarios like school buses that are very low utilization vehicles that might be available at the right time of day to plug in sufficiently big batteries,” said Andy Lubershane, senior vice president of research and strategy at Energy Impact Partners.

He does not expect the practice to take off at scale, however, because demand will not be large enough during overnight hours when most EVs will be plugged in at home. Conversely, when vehicles are in use during the day, owners are more apt to pay for fast charging than to accept a nominal payment to delay charging and commit their batteries’ capacity to a grid operator.

Song agreed that consumers would not be willing to use their cars as storage solutions. However, a stable storage medium is needed for gas stations to effectively incorporate fast-charging terminals.

“They face huge swings in power requirements, forcing them into higher grid cost structures,” she said. “They need a solution that’s connected at the gas station” that allows them to level off their grid draws and store power for use as vehicles need it for fast charging.”

Swaminathan noted that outside the U.S., Asia, and India, the electric grid may not be able to handle V2G tasks and that there would be little opportunity to use it anyway in areas where e-bikes and scooters provide much of the transportation. Ju Li picked up the point, explaining that current EV battery models are not designed to support flowing power back to the grid. “Until this use is considered along with maximizing energy density, we will continue to talk about V2G in terms of potential rather than actuality.”

Hong Li — Professor at the Institute of Physics, Chinese Academy of Sciences (IPCAS) and co-founder of WeLion and HiNa, said V2G tests in China reveal at least five more years are required to commercialize the technology.

“We need automatic driving, parking, and charging and a reduction in the price difference between the electricity consumers pay when they charge and the remuneration they receive when they return power to the grid. “The battery and other technologies are not ready.”

One area in which the technology may be ready for commercialization is material recovery and recycling, which is fortunate because “we probably don’t have enough material to make the number of batteries needed to achieve the goal,” Applebaum said. “We have to look into alternatives.”

Most recycling companies are concentrating on recovering the batteries’ cobalt content,” Ju Li said. “The nickel is not fully recovered, and lithium recycling is less than 20 percent. That’s really deplorable.”

Part of the reason, Swaminathan said, is that recycling takes a hydrometallurgical approach to make the best of today’s market realities.

“If you start with a clean slate, I would love to know how you would design for recovery, not just for performance and cost,” he said. “There is an attempt to mitigate some of those costs but there is also an inability to utilize more efficient supply chains and the global ecosystems which a lot of other industries have adopted.”

Applebaum asserted that mining new materials does not consider the full cost of the externalities involved. She cited the mining inputs, the long-term environmental impact, the cost of labor, and the humanitarian concerns when the required elements are mined in conflict zones or countries with low standards of living and little regard for human rights.

But given the reality, “one of the largest costs of battery recycling today is transportation. That is going to continue to be a barrier, thinking about ROI and sustainability. When recycling is done on a mass scale, we will continue to try to localize it, Veronica Li said. And even if all the materials currently in use could be recycled, it would not come close to meeting the materials needs if the world is to accomplish its environmental goals. Mining, therefore “remains a very relevant part of our future and we need to think about ways to mine more effectively.”

Autonomous vehicles and car-sharing could present a more viable way of reducing the environmental impact, however, Ju Li said.

“It changes how we think about vehicles as a resource,” Applebaum said. “The average American’s vehicle sits parked, either at home or at their place of work something like 90 to 96 percent of the time. How do we increase vehicle utilization whether it’s around charging and vehicle-to-grid or through the sharing economy, so we need fewer resources.”

The panel also discussed hydrogen fuel cells, solid-state batteries, the integration of renewable energy sources and technology into grid/battery systems, and other issues. TDK Ventures will discuss these topics and more as we dive more deeply into the points out Energy Week panelists brought up.

TDK Ventures conceived Energy Week to inspire a new generation of inventors and entrepreneurs, who will help solve the biggest problems in energy we face. Through these discussions with the who’s who in the energy industry, we hope to stimulate long-lasting conversation and engagement in our EX-community that will lead to meaningful actions to help improve and future-proof our world for generations to come.