How European battery manufacturers can secure their competitive advantage with multi-partner data collaboration

European industry has been at the forefront of automotive innovation since the invention of the internal combustion engine. However, the shift to electric mobility has turned the tide. Today, battery costs can account for up to one-third of total vehicle costs. But most of the batteries installed in electric vehicles come from Asia. Oliver Montique, an automotive industry analyst at Fitch Solutions, says almost 90% of the world's large-scale battery factories, on the scale of Tesla's gigafactories, are located in China. And in 2020, Simon Moores of Benchmark Mineral Intelligence found that China builds battery factories almost every week.

This shortfall in battery production seems impossible to make up. But Montique sees a chance: "Battery technologies are evolving every 18 to 24 months. Europe is so far behind on lithium-ion that it would be better putting investment into R&D [...]. That will be Europe’s angle.”

Many other experts agree that the key for making the European battery value chain successful is close collaboration and data sharing. Companies that share data with partners will be better equipped to compete in the future. Collaborative AI can create new cell chemistries and production processes. It can even help develop sophisticated recycling and disposal capabilities.

And yet, while many executives talk about leveraging data for machine learning with partners, most such projects remain unrealized.

In this article, we show battery manufacturers the advantages of secure data sharing. We also give tangible recommendations to realize the full competitive potential of AI.

Europe’s ambitious goals on battery innovation

Manufacturing battery cells in Europe is both a political aim and an economic necessity. A major milestone was achieved in 2017 with the creation of the European Battery Alliance. The goal of the alliance is to create a domestic battery value chain to drive the transition to clean energy. This can deliver decisive competitive advantages and is a unique selling point - further digitizing the value chain end-to-end is a key requirement for that.

Thierry Breton, Commissioner for Internal Market stated: “The batteries value chain plays a strategic role in meeting our ambitions in terms of clean mobility and energy storage. By establishing a complete, decarbonized, and digital battery value chain in Europe, we can give our industry a competitive edge, create much-needed jobs and reduce our unwanted dependencies on third countries – in short, make us more resilient.”

Vice President Maroš Šefčovič, in charge of the European Battery Alliance, said: “By 2025, our actions under the European Battery Alliance will result in an industry robust [enough] to power at least six million electric cars each year. Our success lies in collaboration, with some 300 partnerships between industrial and scientific actors foreseen under this project alone.”

Why the European battery ecosystem needs to leverage AI to seamlessly innovate together

It is in the interest of every company in the European battery value chain to foster a healthier ecosystem. Joint R&D agreements and collaborations on AI are key. These enable companies to perform cutting-edge research on next-generation batteries while sharing the associated risks. They also help build a pipeline for the future performance improvements that are critical in such a competitive industry.

Close data collaborations allow companies to develop much more reliable, more accurate and higher value machine learning models. This significantly improves the success rate of AI initiatives. Ultimately, this leads to reduced costs and increased battery performance. A huge range of applications will benefit from improvements in storage capacity, durability, and other important properties of batteries. These include key applications such as electric vehicles, medical devices and large-scale storage of renewable energy on the grid.

Why are data collaborations unrealized if they can give such an important competitive edge?

Encouraging collaboration between potential competitors is a huge hurdle despite the global economic challenge and threat. There are many roadblocks that prevent companies from collaborating. These include data privacy, regulatory compliance and the desire to protect proprietary technology and trade secrets. Additionally, there is the problem of distrust—companies simply won't share their proprietary data and intellectual property.

How federated machine learning leads to a paradigm shift

Fortunately, modern technology can solve the trust issue. One of the biggest misconceptions in data science projects is the assumption that source data must be physically shared with partners. The answer to this is to use a federated approach. Here, you bring the machine learning models to the data, rather than the data to the model. This way, data scientists can access wider, more diverse datasets than they otherwise could.

Technically, no sensitive data is ever shared within a federated machine learning set-up. All data stays within isolated and secure environments of the organization contributing it. Instead, only computational inputs or aggregate results are shared. This protects both the confidential data and the intellectual property of the machine learning models. As a result, even fierce rivals, for example, cell pack manufacturers, can collaborate.

Federated machine learning is the key to a strong European battery value chain

This distributed and decentralized approach to R&D dramatically improves incentives for all participants. Federated machine learning allows the creation of robust models based on data from multiple partners in the battery value chain. This promises faster data insights, time savings and lower costs without sacrificing privacy or IP.


As we have seen, collaboration is key if the European battery value chain is to remain competitive in the future. However, there are strong disincentives to collaborate on R&D, especially when data sharing is required. Fortunately, federated machine learning already provides a perfect solution that preserves both data privacy and the related IP. Apheris has created an enterprise-grade platform that delivers federated machine learning. Book a demo today to learn how we can help you revolutionize innovation and R&D in the European battery value chain.

Apheris logo