Battery Data Genome: A Path to a Brighter Renewable Battery Future

The Battery Data Genome (BDG) project aims to compile as much technical data about renewable batteries as possible. Similar to the Human Genome Project (HGP), the BDG is led by researchers at the Department of Energy’s Argonne and Idaho Laboratories as well as researchers in Europe. The process would capture data about batteries from battery makers and analyze the data using AI, to allow for faster and more efficient breakthroughs in the renewable battery space, as the global energy sector moves towards a more battery-centric future. 

This call-to-action, as Argonne distinguished fellow and Joint Center for Energy Storage Research Director George Crabtree described on the Argonne website, will collect and house data from every step of the battery lifecycle, from discovery to development to manufacturing and all manner of deployments. The goals are scientific breakthroughs, usable by both the private and public sectors, to make batteries — from small to large scale — more efficient and longer lasting. 

We interviewed Sue Babinec, one of the co-authors of the call-to-action, and a battery scientist and electrochemist at Argonne.

How does the Battery Data Genome compare to the Human Genome Project? 

It is a transformational idea. If you look at the HGP, when you go back twenty years, people said we’re going to decode the body and share this information and it will unleash capabilities to change the world. It’s large, it’s audacious, it’s aspirational, and it’s very difficult to do. In that regard, it’s the same. The HGP started out as different combinations of public and private information, but ended up public. Even though the data is very useful, it spawned many many industries, and so the value that was generated from it is massive compared to the amount of money put into it. 

For batteries, the analogy is that it is a data-intensive activity. The DNA is essentially, what is the battery, and how does it behave when you use it? That is the transformational data. When you have that data, you can better go ahead and predict — if I have a battery with a certain design and I use it in a certain way, this is how it will perform and this is how long it will last. So a product development cycle that normally takes 15 years will take six months to a year. 

The Human Genome is such an important example because when it started, whoever would have guessed that you could take something that cost so much money, and was so precious, and so scientifically superior, and give it away, and still make a ton of money on it? That is an example that shows open sourcing of data leads to creation of tremendous wealth. A lot of people got really rich, but people who had diseases were solved, and crimes were solved, etc. So you have both public good and creation of wealth. 

How difficult is it to get the data? 

To do data science, you have to have a lot of data. We just need to have, let’s say, 10 to 15 percent of the folks who are using these devices, to share that data with others and then we can do data science. You don’t need all the data. Since releasing the paper, we’re actually finding there’s a lot of people who want to share it. 

What is the advantage for companies to share their data with the BDG?

If you’re a for-profit corporation, you want to see people doing research in academia or the national labs in order to do pre-competitive research on your problem, so you would want to release a certain amount of your data to have your problem solved. They can get others to work on their data, too. Also, when the data is old and no longer proprietary, they can share old data. Old data is very very useful, as useful as new data. 

Giving your data to the community does not mean that anybody understands that it came from you. Because data will go through the national labs, it will be anonymized, polluted, there will be aggregation and disaggregation, so you can take proprietary data, give it to us, we can sanitize it, and nobody will know that you contributed. 

Is the BDG a public service project?

It is, in the end, a profit-making and technology-catalyzing paradigm, and it is funded by the Department of Energy out the door, but then it has to be funded by others. 

What is the information you’re looking at? 

A problem that we referred to in the paper is the life of the battery, how many times you can charge it and discharge it. A lot of that work is the current voltage versus time in that battery as it’s charged and discharged hundreds and thousands of times. That’s like the data that we use here at Argonne, a team of electrochemists and data scientists, and we took that data and used machine learning. Without machine learning and AI and data science, if you want to know if a battery will last 2,000 times, you have to cycle it 2,000 times. We can predict out the data to 2,000 cycles with one cycle. 

There’s also data science for when you’re designing materials that go into those batteries. The same principles apply — if you’re trying to design new cathode solid active materials, you would look at density, particle structure, etc. 

Are you also looking at how the raw materials for batteries are sourced? 

The world already knows what the general options are — sodium, potassium — we know what they are, because it’s basic scientific principles. But data science can help you to do a better job of saying, how am I going to get that sodium out of the solution, how am I going to get lithium out of this brine? The data sciences that we’re talking about in the BDG are related to the fact that I have a raw material and I’m going to build a battery, and what’s the best way to do it? It’s not about, how did I get the lithium out of the Earth? 

How many companies have given you data so far?

It’s in the tens. I think a really important distinction is that when we wrote this paper, we were not advocating for people to give up data that they needed to hold secret. So if you have a security job or work for defense, if you need to be secure, we’re not trying to persuade you to give up your data. We’re just saying if you can share your data, let’s make a path forward. 

The BDG is a paradigm. It is not yet a funded operational organization. We have a gap between the writing of this and forming an organization that says, this is the standard we will use. 

How long away is this? 

I wish I knew. What we have right now is a call to action: this is why it’s important, this is what you need to do, this is how you go forward. 

Where will funding come from? 

It will be a private-public partnership. We have a vision for what it looks like. The call to action paper was carefully written by academics and scientists. We have colleagues who are the commercial end and we said, I’m sorry, you shouldn’t be on this paper because that will tarnish the idea. 

Batteries are ubiquitous. They help with climate change, they help to have electric vehicles, they help to get renewable energy, they help us to decarbonize industry. In that respect, batteries touch all parts of everybody’s lives. This is driven by scientists who have a broad view of how to make things better. It really is born out of scientists being good statespersons, frankly.