Kavita Verma, Chief Executive Officer, Maxwell Energy Systems Pvt Ltd (Endurance Group Company)

Kavita Verma is a seasoned business leader in India’s advanced energy and electric mobility ecosystem. She specializes in driving innovation-led growth in battery management systems and energy storage technologies, enabling safer, smarter, and more efficient electric mobility solutions. With a strong focus on scaling deep-tech engineering capabilities and building integrated energy platforms, she brings together technology, product strategy, and execution excellence to accelerate the adoption of next-generation electrification systems across the mobility sector.

In a recent interaction with Priyanka R, Copywriter at siliconindia, Kavita Verma, Chief Executive Officer, Maxwell Energy Systems Pvt Ltd (Endurance Group Company), shared her insights on ‘How Edge Intelligence is Transforming EV Battery & Power Systems into Autonomous Units’.

Electric mobility technology is advancing very fast as vehicles begin to become software-defined and intelligence-based systems. While previously, the concern was about the performance of individual components, now, there is an increasing focus on system intelligence that involves making real-time decisions for improved safety, efficiency, and reliability of operations. This has led to advancements in edge computing and embedded artificial intelligence technologies, which enable battery systems and other vehicle systems to be able to perform processing at the edge of the network (within the vehicle) to improve reaction times. Predictive diagnostics are helping improve battery longevity as well as minimize battery faults.

Real-Time Edge Computing for Software-Defined EVs

Real-time edge computing is not a luxury anymore; it is an essential part of software-defined vehicles. In an electric vehicle, the ability to make decisions based on thermal occurrences, high currents, cell imbalance, or charger communications has to happen without relying on any cloud-based validation, since every millisecond matters.

Battery pack should be designed in such a way that its behavior is smart and autonomous, constantly sensing, understanding, and reacting to events in real time. With edge computing, a number of interventions can be executed quickly, including adjusting charging or discharging speeds, balancing individual cells, separating issues, and ensuring that the system operates normally despite any abnormalities in its thermal processes.

The cloud technology remains useful for fleet analysis and overall optimizations; however, first-level intelligence should take place at the edge in order to ensure proper functioning of battery packs. It looks like the whole industry is moving toward proactive decision-making instead of reactive analysis.

That's why at Maxwell, we include analytics and configurability options directly into our battery management systems. That is how EV systems become truly dependable at scale.

Edge AI-Driven Predictive Diagnostics in EV Systems

The future of EV reliability will shift from being reactive to being predictive.

While conventional battery management systems are mainly responsible for alerting us about events that have already happened, edge AI provides predictive intelligence allowing for the anticipation of potential events. Constant monitoring of such metrics as voltage drifting, impedance changing, temperature differences, usage patterns, and charging patterns allows detecting early signs of degradation long before the situation results in an actual failure.

In a business setting, it is especially valuable since it could lead to problems related to downtime with all the resulting financial losses as well as significant safety concerns. The ability to foresee such problems as faulty batteries, connectors, thermal instability, or charger compatibility will allow avoiding any breakdowns.

Moreover, it could create trust among end users when a self-learning and self-monitoring system would take care not only about the car but also about people's safety while traveling.

With the advent of multi-chemistry solutions like sodium-ion or solid-state batteries, it becomes increasingly relevant to introduce modular and adaptive algorithms which would provide chemistry-, context-, and experience-aware predictive intelligence.

Balancing Performance, Safety, and Battery Life with Edge Intelligence

Modern battery packs' enhanced energy density results in higher vehicle range. However, it comes at a price of lowering the margin for operational errors. It implies that battery management system designers must seek a balance between performance optimization, safety, and battery degradation prevention.

A way forward is through contextual intelligence. Instead of applying rigid rules based on pre-set threshold values, a new generation of BMS systems needs to implement intelligent adaptive control methods that can account for the reasons behind temperature variations, loads, and their consequences for battery aging.

For example, while fast charging might be convenient for users in the short run, it can shorten the service life of lithium-ion cells considerably. Thus, edge intelligence systems must make real-time decisions regarding whether they should focus on performance, longevity, or safety.

This problem is especially relevant for the Indian market, which operates under extremely hot weather conditions. Consequently, there is a need for localized and immediate thermal intelligence, which will allow making accurate decisions.

This concept also applies to next-generation grid-aware BMS systems that leverage machine learning technologies to enable safe operation of vehicles in V2G mode.

Moving Toward a Unified Energy Intelligence Layer in EV Systems

The industry can no longer allow battery management systems, chargers, motor controllers, and power distribution units to operate independently from each other. They must be seen as a cohesive intelligence layer.

The flow of energy is inherently a systemic problem, not an individual component issue. Charging affects battery degradation, motor requirements affect thermal dynamics, and the grid influences charging techniques. Without a comprehensive intelligence approach, optimization becomes inefficient.

In this context, the embedded electronics sector must play its part in this shift. In the future, those companies able to create and sell an integrated ecosystem rather than just a product will succeed.

This is the direction we see in X-in-One solutions, which combine all functionality in a singular architecture: Battery Management System, motor control, DC-DC converter, telematics, and analytics, to name a few. In summary, the focus should not be on creating the best component but the most efficient system possible.

Traditional BMS systems tell you what happened. Edge AI helps you understand what is about to happen.

Key Challenges in Scaling Smart Battery Systems for Mass-Market EVs

The biggest gap is not technology, it is cost-effective scalability.

While mass-market EVs require intelligence, they have to be cost-effective too. The OEMs require intelligent systems which are capable enough to deliver reliability and performance, while being sufficiently modular to avoid complex or excessive engineering.

There are three critical limitations which need special attention: localized manufacturing, software adaptability, and standards maturity.

Firstly, advanced electronics should be produced in proximity to end customers. Secondly, platforms should be adaptable and flexible to the needs of different OEMs, rather than require total customization for each case. Finally, interoperability standards for vehicles, charging stations, and battery storage infrastructure are not fully matured yet. Commercially viable solutions will become available once intelligence becomes a feature of the scalable platform capabilities, rather than a premium add-on feature.

This is why India’s opportunity is significant, we can build not just for adoption, but for leadership, especially when advanced electronics are designed and manufactured locally with scale in mind. Maxwell’s in-house SMT capabilities and India-built systems reflect that direction strongly.

Also Read: Why India's EV Growth Hinges on Grid Readiness and Energy Management

How Edge Autonomy is Changing Power Electronics Design in EV Systems

Power electronics cannot be considered as stand-alone hardware anymore. They should be developed as an integrated solution, with hardware defining the range of capabilities of power electronics and software being responsible for smart use of those capabilities.

Edge computing technology requires all loops, including current sensing, thermal regulation, charger communication, and others, to work in complete synchronization. In turn, it implies the need for co-design solutions, including semiconductors, firmware, control, and analytics.

For example, wide-bandgap devices such as Gallium Nitride and Silicon Carbide allow achieving impressive efficiency results; however, maximum performance can be gained by means of intelligent control techniques based on dynamic adaptation to load and thermal changes.

Another example is related to the usage of non-rare earth motors in EVs. However, their maximum efficiency can be achieved only thanks to embedded intelligence. The design philosophy is shifting from ‘protect the system’ to ‘enable the system to think’. That is the future of autonomous energy systems and it is where the next generation of EV differentiation will come from.

The Shift toward Integrated System Intelligence in Electric Mobility

It is evident that the development path for electric mobility will lead towards an intelligent system that is completely integrated together, and instead of judging its worth based on hardware performance, the efficiency of such a system as a complete unit will be the deciding factor for its value. With the increase in the number of software elements in vehicles, embedding intelligence at the edge is becoming one of the crucial points of difference between various products, allowing for quicker decision-making, increased reliability, and better energy management.

Nevertheless, the move to full autonomy would require more than just technical readiness; it would need the ability of the sector to strike a balance between innovation and the ability to scale, be efficient in terms of costs, and overcome practical barriers for implementation. The localization of manufacturing, flexibility of the platform, and integration at the ecosystem level would be key factors in this regard.

In essence, the direction that EV systems will take in the future will be one of moving from an era of optimizing components to one where intelligence is key, with adaptability, predictability, and energy flow coordination becoming the pillars of success.