E-Mobility Engineering 014 l InoBat Auto dossier l In Conversation: Brandon Fisher l Battery monitoring focus l Supercapacitor applications insight l Green-G ecarry digest l Lithium-sulphur batteries insight l Cell-to-pack batteries focus

Focus | Battery monitoring Acknowledgements The author would like to thank Mike Kultgen at Analog Devices, Nat Edington at Dukosi, Dr Florian Ullrich at InnovationLabs, and Taylor Vogt at Texas Instruments for their help with researching this article. be downloaded at various points, perhaps on an hourly or daily basis, from the BMS. That allows car makers and fleet operators to monitor the performance of the battery pack, and gives visibility of the performance of the cells that is not possible to achieve in the actual vehicle. For example, if there is a problem with a particular pack, the analysis can show any other packs that exhibit the same anomaly, and recall the vehicle before the problem occurs. This can be performed automatically using machine learning algorithms on the data in the digital twin. With more accurate data from the cells, the OV and UV limits can be tighter, and any problems with the current falling outside that band can be identified faster. This reduces the risk of false alarms and inconvenience to the user. A separate digital twin can be set up for each battery pack. This can then be used to track the individual circumstances of the operation of a vehicle’s pack, with all the specific data from charging times and current profiles and how the pack discharges as it is used. Having this data as a digital model allows operators to monitor each cell in the model individually, and highlight any potential problems. This can then be checked out in the real-world version before the issue causes a failure. Conclusion Battery monitoring has an impact on many aspects of the design of an EV. Chips with more accurate current and voltage measurements are enhancing the performance of the battery pack, providing longer range and more reliable operation. Identifying failing cells in a module can trigger the ‘limp home’ modes that stop vehicles failing at the roadside and avoiding catastrophic failures. Local data storage and analysis on the chips are providing more sophisticated monitoring that stays with the cell throughout its lifetime. New wireless connectivity techniques are reducing the weight of the wiring harness and providing more flexibility in extending the size of a pack without having to rely on a modular structure. More accurate data from monitoring all the aspects of the battery pack is also being fed into digital models. The models can be used by the onboard BMSs for comparison with the real-time data to highlight any cells that have a problem, and predict potential problems that might occur. This more detailed data can also be used in the cloud as the starting point for a digital twin. Combining the original r&d model with detailed data from the operation of a specific battery pack – when charging as well as discharging – builds a valuable virtual model. This model can be interrogated in the cloud as an alternative to invasive monitoring, again providing a technique for highlighting potential problems before they occur. ;hin-film sensors allow engineers to measure the state of charge directly and detect any irregular beha]iour (Courtesy of 0nno]ation3abs) 40 Summer 2022 | E-Mobility Engineering