As India’s quick commerce and urban mobility sectors expand, keeping EV fleets running round the clock requires more than just batteries. It demands intelligence.
Battery performance remains one of the biggest challenges in the EV industry. The efficiency, lifespan, and charging capabilities of batteries determine the overall success of EV adoption.
AI-driven Battery Management Systems (BMS) analyse real-time battery data, including temperature, charge cycles, voltage levels, and driving patterns, to optimise energy consumption and prevent degradation.
Reflecting the growing demand for AI-integrated battery solutions. By employing predictive analytics, AI can prevent battery failures, enhance safety, and extend battery lifespan by up to 30%, reducing the need for premature replacements.
EMO Energy, which consults companies through its AI-powered battery management system, focuses on two primary areas: energy management and battery health prediction.
Sheetanshu Tyagi, CEO of EMO Energy, in conversation with AIM, stated that EMO explores solar energy, battery storage, and grid usage for its charging stations through a comprehensive system.
The Bengaluru-based startup company utilises Long Short-Term Memory (LSTM) models to forecast energy demand at various charging stations. It strategically positions them near dark stores to address specific energy needs. By combining vehicle data and environmental factors, such as traffic density, EMO can determine the optimal locations and quantities for charging stations.
The company claims their solution enhances battery life prediction by utilising essential data and LSTM models, forecasting battery life up to two years based on data from 1,000 batteries. It also uses Convolutional Neural Network (CNN) techniques to monitor heat concentrations in battery packs for optimal performance. Additionally, EMO says it helps dark store owners and companies like BigBasket, Blinkit and Zepto, which have partnered with them, manage energy and mobility, keeping costs low.
Energy Management and Cost Benefits Across Cities
“Energy is the only input cost that can be systematically optimised over time,” Tyagi said. “Fuel prices, vehicle rents, labour, and real estate continue to rise annually, whereas EMO’s vertically integrated ecosystem is designed to flatten and even reduce energy costs over time.”
In Tier 1 cities like Bangalore and Delhi, the baseline vehicle costs, such as rent, fossil fuel, and workforce, are higher, which immediately gives EMO a cost advantage on the EV front. Transitioning to its energy ecosystem results in a faster payback period and greater operating leverage, the company said.
In Tier 2 cities like Mysore and Indore, although the same percentage savings apply, the absolute monetary gain per rider is lower. However, the network effect of the platform still leads to increased delivery efficiency, meaning more productive kilometres can be covered per unit of energy consumed, resulting in significantly reduced idle time.
India’s ability to operate in diverse environments is a significant strength, especially in cities like Bangalore and Gurgaon, which have distinct characteristics.
Factors such as energy storage and power utilisation also vary. Gurgaon’s heating, ventilation, and air conditioning systems require higher energy and exhibit distinct patterns of solar energy. These variables significantly influence training models and determine outcomes for both riders and operations.
However, Tyagi said, “The growth of quick commerce is evident, as I’ve seen the number of dark stores being set up. The focus is that by 2030, there should be almost 40,000 dark stores in India.”
Rider Deployment
Previously, quick commerce companies needed a dense cluster of riders per hub to meet unpredictable demand patterns, with each rider typically covering only 70 to 80 km per day. With EMO’s energy ecosystem, riders are now consistently achieving an average of 140–160 kilometres per day, Tyagi added.
EMO’s partnered companies, as a result, could reduce the workforce overheads while enhancing their delivery service level agreements. The exact number of delivery kilometres can now be covered with nearly half the number of riders, the CEO said.
Fewer riders per hub also lead to lower energy usage, reduced idling, less charging downtime, and more effective utilisation of each vehicle and battery asset. Consequently, partners such as Blinkit, Zepto, and BigBasket have been able to reduce rider deployment while maintaining or improving fulfillment rates and lowering the overall delivery cost per order, according to the company’s data.
The company’s analysis works on three key parameters: voltage, current, and temperature. These factors are critical for calculating the internal resistance of various cells in a battery pack, which can contain hundreds of individual cells, Tyagi said. By using a temperature heat map and measuring the voltage drop, an internal resistance map is created, providing insights into energy loss at the anode and cathode.
The team is currently focused on basic data, such as riders’ generalised locations and environmental temperatures. They plan to incorporate terrain-based data, such as gradability inputs, in the future to enhance range predictions for various routes.
“We’ve seen up to a 30% difference in EV range depending on how and where a rider operates, so we’re building dynamic rider scores that factor in behaviour, charging patterns, and environment to shape financing and insurance models for the future,” Tyagi highlighted.
Energy Efficiency
Tyagi highlighted two points. First, the five established hubs have developed operational patterns for dark stores, which vary by industry. While the frequency of restaurant operations is decreasing, dark stores continue to be a rapidly growing segment. The company’s analyses of electricity patterns revealed peak energy loads in the afternoons due to air conditioning use.
He said that this insight allows for optimising battery charging based on grid-level data, seasonal variations, and location-specific factors. A machine learning model predicts the best times for solar charging, grid usage, and battery deployment for each hub.
Secondly, the team is focused on scaling their infrastructure to improve battery and charger systems, integrating IoT devices along the way. At its core is a comprehensive battery lifecycle management platform that is being integrated with various OEMs.
Investors are also increasingly betting on integrated infra-tech startups. When asked what’s driving the shift from OEMs to the tech infrastructure place, Tyagi said companies across various sectors, including AI and data centres, are increasingly interested in integrating their entire infrastructure to ensure reliability.
By being closely tied to the end user, the company can better manage revenue cycles and scale more effectively. He emphasises that their most significant barrier to scaling is often self-imposed, aside from external dependencies.
The post This Bengaluru-Based AI-Led Battery Startup Powers India’s EV Future appeared first on Analytics India Magazine.
