Eatron’s intelligent software layer unlocks the full potential of batteries

Electric vehicle batteries are designed to operate within a predefined window. While this helps to ensure consistent performance and guard against the effects of degradation, it does mean that as much as 5-10% of a battery’s capacity is unusable. But what if we could unlock this hidden resource?

Eatron Technologies’ Intelligent Software Layer (ISL), which can be deployed even on top of an existing battery management system (BMS), has the potential to do just that, leading to better performance, an extended lifetime, and increased safety.

Better performance

The secret lies in Eatron’s proprietary algorithms, the most advanced on the market, and with best-in-class accuracy. Our technology achieves a consistent SoC (state of charge) accuracy of ±1%, compared to an industry benchmark figure of around ±3%. Our SoH (state of health) algorithms are similarly precise, maintaining a ±2% accuracy throughout the lifetime of the battery, in contrast to an industry average figure of ±5%. With a clearer and more accurate picture of the state of the battery, it becomes possible to extract the maximum performance from it without exceeding the safety limits.

Extended lifetime

Our patented approach to predicting the battery’s Remaining Useful Life (RUL) uses a blend of our innate understanding of battery chemistries and technologies combined with our experience of machine learning and AI. Using hybrid physics  & AI based models, our methodology applies techniques such as feature extraction to transform the vast quantities of data coming from the battery into smaller streams that focus on the factors we know are important. This helps to reduce the volume of inputs, simplifies the processing required, and vastly speeds up the prediction process.

By monitoring the health of the battery over its entire lifetime, it’s possible to react to changes within the pack and potentially even work around them, perhaps by deploying alternative usage profiles or charging strategies. As a result, this more accurate picture of a battery’s health allows us to safely extend its operational life, in some cases by as much as 25%.

Of course, once a battery pack eventually reaches its minimum performance level, it is ready to be repurposed into its second life, accompanied by a tailor-made assessment of its condition as part of its battery passport. Passports will become mandatory for all EVs sold in the EU from 2027, and Eatron’s offerings are fully compliant with all forthcoming EU battery regulations.

Increased safety

Detecting fault conditions such as lithium plating is critical to maintaining battery safety, and our AI diagnostics are able to predict cell failures weeks before they occur, with 90% accuracy, zero false positives, and across all major battery chemistries.

Lithium plating is caused by lithium deposits that collect around the anode, and most often occurs when charging at low temperatures or at high currents such as during fast charging. Over time, these deposits reduce the capacity of the battery and, if left unchecked, can eventually cause a short circuit that can induce thermal runaway.

As well as avoiding the types of catastrophic failures that have been well publicised, this ability to predict critical failures in advance also avoids scenarios where a vehicle unexpectedly runs out of power, itself a key safety concern.

As an intelligent software layer, all of this functionality can be deployed quickly and without the need for huge upfront investment by the customer.

Eatron’s algorithms can be deployed either in the edge and/or in the cloud, while a hybrid mix of the two delivers the maximum benefit in terms of accuracy and efficiency. For example, performing the feature extraction in the edge allows us to analyse the high-frequency data on the device and only transmit the most pertinent data to the cloud, massively reducing bandwidth consumption. There it can be analysed further in conjunction with data received from thousands of other vehicles to arrive at highly reliable insights gained with maximum confidence yet low computational effort.

However, even an existing fixed BMS can still benefit, by deploying Eatron’s algorithms within the cloud to add an additional layer of functionality that would not be possible otherwise. And, as we continue to invest in our cloud-based technology, our clients benefit from new enhancements that increase accuracy and efficiency still further.

Alternatively, it can be embedded into a complete hardware solution, and Eatron already offers a reference design thanks to its connections with established semiconductor partners.

Whichever models works best for your application, Eatron’s intelligent software layer will continue optimising the battery’s performance throughout its life, from the day it leaves the factory, to the very last mile.

Amedeo Bianchimano, CCO at Eatron Technologies.

At Eatron we develop intelligent Battery Management Software solutions that can be deployed on both the edge and the cloud. If you are interested in learning more about these click on ‘Talk to an Expert’ and fill in the contact form or email directly.