Nio, Monolith join forces to offer real-time AI-enabled EV battery testing

Chinese electric vehicles (EV) manufacturer Nio is collaborating with a UK AI software startup, Monolith, to enhance electric car battery packs using real-time testing.

The collaboration enables the automaker to assess new batteries by analyzing their health, performance, and durability every time they are swapped. Monolith’s AI algorithm processes the data, providing real-time insights for battery improvements.

The partnership aims to improve battery performance by leveraging data from Nio’s battery-swapping service in Europe. The project’s testing will soon begin in five European markets.

“ Monolith’s deep-learning algorithms allow for automatic detection of battery issues, such as spontaneous discharge and thermal runaway, making it easier to analyse complex real-world data quickly. This capability not only saves time and resources but also enhances battery safetym,” said Richard Ahlfeld, CEO and Founder of Monolith, in a statement.

Efficient battery testing

Electrification is transforming the automotive industry, driving significant investment in developing longer-lasting, faster-charging, and cost-efficient batteries.

Traditional simulations fail to accurately model battery performance, forcing OEMs and cell manufacturers to rely on physical testing, which increases costs and delays. A new approach using AI and machine learning is essential for staying competitive.

Monolith’s AI models, built from test data, offer a simpler, more accurate alternative to traditional Error Correction Models (ECM) and Pseudo-Two-Dimensional (P2D) models. With Monolith’s interactive tools, users can analyze battery designs, predict how batteries will perform under various conditions, and optimize key parameters.

According to the firm, the platform highlights which design elements and operating conditions have the most significant impact on battery performance and lifespan.

Machine learning models in Monolith predict test outcomes across multiple conditions, allowing users to adjust inputs in real time and see the effects instantly. Active learning techniques optimize test plans, reducing unnecessary steps while maintaining coverage.

“Using valuable and sometimes limited engineering test data, the AI software makes instant predictions and enables engineers to identify areas where optimisation and development are required, without the extensive need for repetitive, time-consuming physical tests,” said Monolith, in a statement .

Advanced battery insights

Using NIO’s battery swap technology, Monolith will now have a rare chance to learn new things about how batteries are actually used regularly.

It will track performance based on field-generated data using Monolith’s Anomaly Detector AI engine. These insights will be incorporated into additional verification efforts and serve as a foundation for comparing test-bench outcomes.

According to the companies, based on intricate system linkages, NIO and Monolith engineers can identify anomalies in cross-channel findings at a rate never before possible with an Anomaly Detector.

Monolith’s self-learning algorithms now enable quicker and more efficient detection of battery test data anomalies. By automating the inspection of raw test data, the system identifies potential errors or abnormalities across hundreds of test channels.

“NIO’s partnership with Monolith exemplifies our commitment to delivering a premium user experience. 98 percent of our users across the five key European markets opt for Battery as a Service (BaaS). BaaS allows NIO users in Europe to flexibly charge or swap their batteries as needed, enhancing convenience and cost efficiency,” said Frank Kindermann, Head of Battery System Europe at NIO, in a statement.

Monolith says the company is also navigating regulatory requirements to expand into China as well. This move could allow Monolith to collaborate with other Chinese automakers.