Batteries are not just large chemical systems capable of producing electrical energy. The BMS (Battery Management System, for its acronym in English) is the management system that allows the batteries to work at all times within the most appropriate parameters. However, current systems they are very limited because it cannot learn from its past or from the behavior that other drummers have had in different circumstances. That’s why, a BMS based on data stored in the cloud and managed by the appropriate algorithms could help get more out of electric vehicle batteries.
The BMS is an electronic system that manages the recharging and discharging of a battery. Among other things, it is responsible for preventing it from working outside its safe operating area by collecting data on its status, controlling the individual voltage of each cell and balancing it when necessary. It is an intelligent software whose algorithms may vary from one manufacturer to another and which is key to extending battery life. The BMS accompanies a communication system with the outside which is commonly known as the data BUS.
Dr. Kai-Philipp Kairies, Executive Director of ACCURE Battery Intelligence, has been a researcher, consultant and academic in the battery industry for a decade, conducting his work in places as diverse as Germany, Singapore and California. Thanks to them, he has accumulated extensive experience in the field of technological innovations in the field of energy storage.
The startup he currently runs drives research and helps other companies understand and improve the safety and longevity of their batteries through advanced data analytics. In an interview conducted by BatteryTechnology, describes the scenario in which your field of work finds itself and explains the role it will play in the future data-driven battery technology.
Lithium-ion batteries play an essential role in the new mobility because they are responsible for storing energy, which must be of renewable origin and power the vehicles in which we travel. The common industry goal of achieving net-zero emissions goes hand in hand with improving battery safety, performance and lifespan.
From the environmental point of view, “we need to use the batteries we have today more efficiently,” says Kairies. This need is also applicable in terms of scarcity of raw materials and saving of resources. The battery data they can indicate how charging cycles and other usage factors affect its lifespan and performance, revealing how to improve it. “We must spread the word that safe, reliable and sustainable batteries are possible and are the backbone of our energy future.”
The BMS monitors all cells within a battery pack balancing them to limit the divergence between them. It also calculates the state of charge (SOC), with which the autonomy of an electric car is determined (or the time left to use a laptop). Finally, it includes basic safety features such as protection against overheating or deep discharge, which try to prevent catastrophic battery failure.
However, the BMS does not recognize when cells behave slightly irregularly and, therefore, cannot inform the operator to warn him of a danger. The reason this happens is your data access limitation. The BMS cannot see its past or the behavior of other similar battery packs, which would provide a baseline for advanced diagnostics. This barrier restricts the ability to predict the health status of the battery.
Therefore, although the BMS generates all the data needed to perform advanced analysis, you can’t do it locally. “That’s why to operate batteries effectively and successfully prevent fires and unexpected breakdowns, an extra layer of safety is needed,” explains Kairies.
Batteries generate large amounts of data during their life cycle. Most of this data is not currently used because the BMS is not designed to “translate” this data into actionable information, forecasts, and predictions. With a “translation program like ACCURE Cloud, we can apply our algorithms and database knowledge of the behavior of millions of battery cells to provide intelligence to battery operators that goes far beyond what BMS can do.” The program makes really transparent what happens inside a battery, without the need for expensive tests laboratory which also require a lot of time. No additional hardware required or make any modifications to the battery system. The “intelligence” is based purely on operational data that is already available.
“Once we have the BMS data in our ACCURE cloud, it is filtered so that it is available in a structured way.” The objective of this data cleaning is to manage its complexity, unify the formats and correct measurement errors. “We use machine learning on the one hand and the knowledge of our battery expert on the other hand to prepare the data.”
Subsequently, a virtual model of the battery is built (what some call a digital twin). By extracting features such as voltage, lithium loss, and complex impedances from field data, these digital twins allow, among other things, to detect anomalies with the potential to cause incidents such as battery fires, predict future performance and aging and optimize operation.
“We use these algorithms in a wide variety of applications, from micromobility and electric buses to sports cars, ships and large-scale energy storage facilities. Our advanced analytics also help determine the value of retired batteries to find suitable second-life applications for them.”
The need for batteries with higher energy densities and longer lifespans drives innovations on all fronts: cell chemistry, packaging design, BMS, and cloud analytics. “Personally, I don’t think BMS needs to be much better than it is today. It would be great to have some additional data, like ultrasound or EIS online, but it doesn’t look like it’s going to be commercially viable anytime soon. In the meantime, I think cloud analytics is a fantastic solution to apply right now.”
