We are at a cusp of transition from combustion based vehicles to batteries operated vehicles. With all new elements making an electric Vehicle (EV), monitoring the data helps not in better optimization of the EV performance but also in improving the succeeding vehicle versions. EV data mostly comes from the onboard Electronic Control Unit (ECUs) and Battery Management Systems (BMSs). The data generated by EVs, which comes from sources that vary from their subsystems like BMS, Motors and controllers, to drivers, charging stations, and infrastructure, constitute the big data of EVs. They also carry dozens of sensors that provide data including user driving behaviours, battery security via the BMS, and grid charge management via charging stations. In addition to the data directly collected from EVs, drivers can voluntarily share information about their driving patterns and charging habits. Trip information including start and end times of journeys, connect and disconnect times of chargers, and the battery state of charge (SOC) can easily be collected. Advanced systems can record details like how a driver accelerates or breaks.
An important parameter for EV performance is the driving range. In India, the market adoption of EVs is low due to range anxiety, which is the worry that the EV battery will run out of power before the passenger reaches the destination or a suitable charging point. Big data is frequently used to estimate the driving range, which is an efficient way to diminish the range anxiety
All these various kinds of data can be used for decision making through data analytics tools. Big data analytics have been useful for EV integration in a variety of ways like optimised charging, battery management, and EV status tracking. Mobility needs of drivers can usually be captured with data tracking devices, which in turn helps to understand energy consumption profiles of drivers. Data analytics is also important for utilities to control charging. The utility will eventually decide which services are needed by the EVs via analysing its daily demand data stream. In addition to aiding planning for utilities, decision-making tools need to account for user convenience like each EV having satisfactory SOC at morning departure, as well as an emergency driving range.
Once this vast amount of data is analysed, they can be used to develop policies for siting charging stations, developing smart charging algorithms, solving energy efficiency issues, evaluating the capacity of power distribution systems to handle extra charging loads, and finally, determining the market value for the services provided by electric vehicles, for example, vehicle-to-grid opportunities.
Today, the expected life of the battery is largely unknown, and the key to unlocking the mystery of battery life lies in the data. Battery degradation is a natural process that permanently reduces the amount of energy a battery can store. Battery life can be affected by usage, charging patterns and the environment in which they operate, among other things. Since batteries are a dominant component of the cost of EVs, their optimal usage and proper maintenance can be the key for successful adoption of EV. The underlying, core potential of battery data, when leveraged with artificial intelligence and machine learning can help accurately determine, predict, and exceptionally improve battery life.
Artificial Intelligence (AI) is set to disrupt the battery technology space, by combining the power of predictive intelligence and data analytics to achieve high battery efficiency, and operational reliability. SOC of EV battery is a key parameter for most charging and discharging decisions. BMS logs show SOC information and how an EV battery is performing. Malfunctioning battery cells, and heating and cooling details can be observed via these logs. Smart battery solutions provide extensive system diagnostics like accurate cell voltage, state of charge, temperature monitoring, cell balancing, real-time with the help of IoT and analytics. Machine learning (ML) brings a layer of intelligence, after gathering and monitoring extensive data on battery life, performance, state of charge, stress from rapid acceleration and deceleration, temperature, number of charge cycles, etc. that are stored on the cloud. The power of predictive intelligence is used to predict battery life, safety threats, identify potential breakdown and their causes, fix delays/errors even before they arise. Machine learning makes sense of battery data, brings visibility into battery health and performance, derives valuable insights, and suggests actions that can significantly improve the battery life, reduce downtime and the overall ownership cost.
Esmito has developed a battery analytics tool to assess how batteries have been performing and predict its residual life. We have analysed the battery health of various OEMs across India and gained insights on the real-world conditions that influence the battery health of EVs. The tool also generates insightful reports on battery usage and safety alerts for respective stakeholders in the ecosystem. It will help ensure cost optimisations and zero downtime, and essentially accelerate the transition of businesses to an all-electric future. Data and associated analytics performed on the aggregated data shall play a very important role in all future modes of transportations.
