Ionworks unlocks PyBaMM for industry users with professional user interfaces

Ionworks is a cutting-edge battery software modelling start-up focusing on bringing the benefits of the Python Battery Mathematical Modelling (PyBaMM) framework to a wider audience. Its software is important for addressing industry challenges such as modelling battery degradation diagnostics in real world scenarios and shortening prototype development cycles. Following the completion of a Faraday Institution Entrepreneurial Fellowship, in which the company developed professional user interfaces, we caught up with Dr Tom Tranter, Co-Founder and CTO of Ionworks, to discuss their progress in the last year.

What is PyBaMM?

PyBaMM is an open-source platform written in Python that enables efficient simulations of battery performance and aging, accelerating battery design and innovation. As an accurate, physics-based modelling platform that uses mathematical equations that represent the underlying physical processes within a battery, it can help users predict battery behaviour and understand barriers to performance improvements.

Tom explains: “It can take years to test battery degradation profiles in the lab. And there you run the batteries in really harsh conditions that don’t represent their experience in real-world use cases. Modelling can help battery developers make informed decisions and explore the design space more effectively, reducing prototype costs and shortening battery development cycles in the process. Modelling can also help translate insights from accelerated ageing in the lab to real-world applications.

Because PyBaMM is open source, students and academics often become familiar with it as their first battery modelling tool and can directly see and learn what’s going on by looking at the code. It’s now grown to be one of the most important research tools—especially for battery R&D problems. It also has enormous potential for control problems where data can be collected from the field and insights generated through digital twins, which you can’t really do with any commercial battery package out there.”

Growth Under the Faraday Institution Entrepreneurial Fellowship

Cutting edge physics-based battery models have the potential to bring considerable benefits to organisations across the battery space. But the user interface for PyBaMM and the complex mathematics was holding back its adoption by teams adjacent to R&D, particularly SMEs that don’t have in-house development teams with a modelling or coding background.

The aim of the Entrepreneurial Fellowship was to develop professional interfaces—a prototype software-as-a-service (SaaS) platform for PyBaMM – making the code more attractive and usable for a wider audience – particularly in industry.

“We were awarded the fellowship in the early stages of forming the company. It gave us time to develop our business plan, pay our own salaries, do some consulting work, as well as hiring someone full time. That took time as we hadn’t got involved with full stack development or managing a SaaS platform before. Had we not received the funding we would have had to focus on a lot more consulting projects, and it would have taken us much longer to develop our new platforms,” said Tom.

Ionworks has developed two prototype products. Ionworks Lab, which is designed for people who are familiar with PyBaMM and have the core coding knowledge to change the models, but who might struggle to create models from their data quickly and effectively. The second, Ionworks Studio, is for people who don’t have any coding experience and want to explore what the models can tell them for different design scenarios and use cases.

“Ionworks Studio is a departure from what PyBaMM was traditionally focused on. It has been tailored for the mass market,” says Tom.

Over the course of the fellowship, Ionworks has secured high-value 6-12 month contracts with several companies. Clients include material manufacturers, fast-charging companies, and companies developing electric vertical take-off and landing aircraft. The consultancy side of the business also led into IP and product development, and the tools developed during those projects are now licensable.

Despite the progress, the last year hasn’t been without its challenges. “We’re all new to this and we’ve all got quite a similar background. None of us have ever founded a company before. And none of us have an MBA or any formal training in business. So, it’s all been a bit of a learning process. However, the conversations we have with customers usually start with the technical users of our tools and our reputation for quality and openness has set us in good stead.”

Giving Back into the PyBaMM Community

PyBaMM now has a community of thousands of users globally, and almost 100 contributors who donate their time to furthering the development of PyBaMM. Tom explains their company’s relationship with PyBaMM, “As a very young company, Ionworks doesn’t yet financially contribute to PyBaMM, but we do contribute a lot of time into its maintenance. We are very active on GitHub discussions and the slack channel, helping users solve problems. We encourage our customers to allow us to contribute new physical models, algorithms, solvers and infrastructure improvements back to the open source. We are also sponsoring undergraduate and post-graduate internships this year, which will be entirely focused on the open source.

Ionworks wants to stay close to the PyBaMM community, to work alongside it, to contribute to its independent steering committee, and to provide commercial platforms that some may want and need, but not replace existing PyBaMM tools liked by traditional users.

Looking to the Future

Now that the fellowship has ended, Tom said they have “another year ahead where we’re going to be focusing on execution. We really want to get as much traction as possible with the products that we’ve already developed to generate revenue and have a bigger impact. R&D is the start of our journey but physics-based models coupled with data management, hardware-in-the loop, advanced battery intelligence and optimal control strategies are all potential future avenues we could go down.”

Case study published in May 2024.