How AI and Smart Batteries Are Making Electric Mobility Safe, Reliable, and Affordable

✍️ Opinions

Sanvi Barjatya

10 Jun 2025 — 5 min read

How AI and Smart Batteries Are Making Electric Mobility Safe, Reliable, and Affordable, Arjun Sinha Roy, Co-founder of iRasus Technologies

This Article has been contributed by Arjun Sinha Roy, Co-founder of iRasus Technologies.

One of the key challenges of any tectonic Shift in Technology is the underlying fear of the same. The early days of Aviation, Railways, and Automobiles were often riddled with technology that was unpredictable, hindering widespread adoption. Technology changes always focus on making systems more efficient and redundant, thereby making all systems more foolproof, safe, and reliable.

The early days of adopting electric vehicles and battery storage have also brought forward numerous issues. Some of the key challenges across the industry are as follows.

Safety – Issues surrounding thermal runways in the early days had created a lot of doubts in the minds of electric vehicle users. This, coupled with issues around Range Anxiety (the inability to predict how long the battery would last accurately), has all contributed to the initial misgivings about Electric Vehicles
Reliability – Battery storage-based mobility systems require reliability. Have you ever seen the old Bollywood movies from the 60s or 70s, where the heroine is driving the car and the car stalls, and the hero appears suddenly out of the blue, opens the bonnet, repairs and becomes the knight in shining armour? Well, modern IC cars will disappoint today’s heroes. Today, technology has advanced to the point where software controls various functions and features with multiple levels of redundancy, ensuring reliability and efficiency.
Affordability – Systems and engineering at scale do not fail, thanks to modern quality assurance and various reliability checks, as well as redundancy. The idea behind modern systems is to utilise available technology to reduce errors to a negligible level, reuse systems to minimise costs, and thereby make the product affordable to all.

How is iRasus helping achieve the above for the EV industry? iRasus is unique in taking a software and AI-first approach. It works with all kinds of Systems and Hardware to make it simple for all stakeholders, ensuring interoperability, ease of use, and thereby reducing friction for all stakeholders within the system.

First and foremost, iRasus works with all stakeholders in the ecosystem. Our customers include Battery Manufacturers, OEMs, Fleet Operators and Finance Companies (who finance the Vehicles).

The first approach is to ensure that data is available from the core systems and their adjacencies, i.e., Battery, Chargers, and Motors.

The battery accounts for 50% of the value of an electric vehicle and has the most profound impact on multiple aspects of the vehicle's performance. A thorough understanding of the Battery helps in solving the fundamental problems facing all fleet users.

The full benefit of the above models would be enabled when the entire Data Acquisition, Data Processing and Data Analysis Layers are enabled. Some key points for enabling the use of AI and Intelligence in Batteries are as follows.

Connected First

The first step is to have a connected First approach. All vehicles and Batteries today are smart first and connected first. A Smart Battery Management System (BMS) ensures that data from the Batteries is available every second. Hundreds or thousands of data points are streamed per second, depending on the vehicle's size and the configuration of the available batteries. Our platform decodes the streams of data at high frequency, ensuring every single byte of data from the Battery and Adjacent systems is available for analysis.

If the system is not connected, no amount of edge-level redundancy is enough to monitor actual on-field performance. No amount of engineering can replicate actual on-field scenarios and what we call edge or corner cases —i.e., scenarios that are not possible for human designers to envision.

Building Charts / Patterns

The data received needs to be available in the form of various Charts and Dashboards. These charts and dashboards serve as the foundation for real-time analysis and decision-making. One of the basic charts that are built involves creating a fundamental analysis that plots various charging and discharging parameters, such as voltage, Current, and Cell Temperature. Each of these parameters, along with its corresponding chart, helps the system identify outliers. Each outlier helps in Operations planning. Imagine an electric Bus fleet operator now able to examine each outlier and determine the outcome of each one through proactive planning and maintenance.

Fault Management and Fault Prediction

Battery Management System-generated faults and Algorithmic faults are crucial to understanding what is going on in a Battery-Operated System. The BMS continually generates various types of Fault alerts that are preconfigured and determined by the BMS. These need to be logged and categorised in terms of their severity.

Faults with higher Severity could trigger a standard Operating Procedure (SOP) that focuses on addressing the issues before they become Catastrophic failure. One of the simplest fault codes is over-voltage, which, if consistently observed, could indicate overcharging, a faulty battery management system (BMS) reading, or Wiring/Installation issues.

iRasus also works around developing System Generated Faults. These are faults which. Search for Patterns in the data to determine a potential fault.

Find the root cause of faults through correlation of different input data (something similar to a multivariate attribution model).

The above processes enable the use of AI across various battery storage applications.

AI applications in battery storage from reactive to proactive

With today’s AI and machine learning models for battery data, the above applications can be developed to drive electric vehicles, offering an envelope of operations that is Safe, Reliable, and Affordable. It’s only a matter of time before more and more Systems adopt Battery storage, and the need for AI-based models will be felt to use real-time data and make these Systems SAFE, RELIABLE and AFFORDABLE.