How Vehicle-Cloud Computing–Based Smart Thermal Management Improves New Energy Vehicle Range

It’s a frustratingly common story for electric vehicle owners: The range advertised when you bought the car looks fantastic, but the distance you can actually travel on a single charge often falls disappointingly short. This is especially true in the winter or under difficult driving conditions.

During the harsh cold of a northern winter, for instance, a significant amount of energy must be diverted simply to heat the battery pack and keep it operational. That's power that never makes it to the wheels, resulting in a direct and dramatic drop in your real-world driving range.

How Does the Thermal Management System Affect New Energy Vehicle Range?

To solve the problems of range and performance caused by temperature fluctuations, every modern electric vehicle is equipped with a sophisticated thermal management system. Think of it as the EV's master thermostat. Its job is to keep all the critical components—the battery, the motor, the power electronics—within their optimal temperature range to ensure peak performance, while also maintaining a comfortable cabin for the driver and passengers. This meticulous temperature control has a direct impact on the vehicle's range, stability, and overall safety.

Here, however, EVs face a fundamental challenge that gasoline cars do not: with no engine waste heat to recycle for warmth, they must rely on the battery for everything. This creates an unavoidable trade-off: every time the thermal management system works to regulate temperature, it consumes precious battery power.

The data is striking. The thermal management system is responsible for a staggering 15-20% of an EV's total energy consumption, second only to the drive motor itself. It’s like running the central air conditioning in a massive office building; while the environment is perfectly comfortable, the energy bill is enormous.

Put simply, any energy saved by the thermal management system is energy that can be used to extend driving range.

The critical question, then, is: how can we effectively minimize this energy drain?

Why Traditional Thermal Management Systems Fall Short of Intelligence？

To solve the problem, we first have to understand the core technical bottleneck in the thermal management systems used in most EVs today.

The majority of these systems still rely on simple, rigid control logic based on fixed thresholds. For example, the system might automatically switch on the cooling circuit whenever the battery temperature exceeds 35°C, or start heating it when it drops below another set point. This "one-size-fits-all" approach is simply not effective enough for the complex and dynamic nature of real-world driving.

These conventional systems are unable to intelligently and precisely adjust by synthesizing multiple factors at once—like ambient temperature, individual driving habits, upcoming road conditions, and the battery's real-time state of health.

A perfect example is a short trip on a cool day. The system might "over-condition" the battery, triggering frequent heating and cooling cycles that needlessly waste power. According to our measurements, this kind of thermal inefficiency can consume an extra 1.2 to 1.5 kWh of energy every 100 kilometers.

This makes the intelligent upgrade of the thermal management system a critical breakthrough for any automaker looking to maximize vehicle range and significantly improve the ownership experience.

Industry Challenge: Choosing the Right Path for Intelligent Thermal Management — Cloud-Based Inference or On-Vehicle Computing?

It's now widely accepted that the future lies in intelligent thermal management systems that can dynamically adjust their control strategies based on real-time driving conditions. Leading automakers are already beginning to build these predictive models, but they immediately face a critical question about the underlying technical architecture.

Currently, two dominant approaches are being considered:

1. The Centralized Cloud Inference Model

In this scenario, a powerful model running in the cloud makes all the real-time thermal control decisions, sending commands down to the vehicle, which simply executes them. However, this architecture is fraught with challenges, including regulatory and data compliance hurdles, vulnerability to network instability, and significant, perpetual operational costs for data transmission and cloud computing.

2. The Vehicle-Cloud Collaborative Model

Here, the cloud's role is focused on building and training sophisticated models. Once a model is mature, an optimized version is deployed directly to the vehicle, which then uses its own onboard computing power and local data to make real-time control decisions. While this is arguably the more robust approach, its implementation is a major hurdle. The industry currently lacks mature toolchains, a complete and proven methodology, and deep practical experience, forcing any company that chooses this path to absorb enormous R&D investment and trial-and-error costs.

The EXD Vehicle-Cloud Collaborative Computing Solution: Achieving 5–7% Energy Efficiency Improvement

To solve this architectural dilemma, we at EXCEEDDATA (EXD) have developed a complete On-Vehicle Self-Learning Smart Thermal Management solution, built upon our proven Vehicle-Cloud Data Foundation.

Our approach perfects the vehicle-cloud collaborative model: the cloud is dedicated to building, training, and optimizing the AI models, while the vehicle itself executes inference locally, delivering real-time control strategies that are perfectly matched to its immediate operating conditions.

The results are tangible and significant. In terms of concrete energy savings, our solution enables the vehicle to achieve a 5% to 7% reduction in overall thermal energy consumption. This translates directly into a meaningful increase in real-world driving range and a dramatically improved ownership experience.

Most importantly, we provide a clear and proven path to implementation. Our solution is built on EXD's mature vehicle-cloud collaborative architecture, allowing automakers to bypass the costly and time-consuming trial-and-error phase that plagues in-house development. By leveraging our comprehensive vehicle-cloud data foundation toolchain, we streamline the entire process—from low-code model building and one-click deployment to continuous iteration.

This not only makes the superior technical architecture accessible—it makes development on it fast, efficient, and cost-effective.

To be more specific, the solution is implemented in three distinct phases:

Phase 1: Cloud-Side Model Construction

The process begins in the cloud. Using the EXD Vehicle-Cloud Data Foundation, we collect a broader and more precise set of on-vehicle data. This rich dataset is used to build an intelligent model composed of four core modules: Trip Prediction, Driving Style Recognition, Vehicle Energy Flow Modeling, and a Planning & Decision-Making engine. Once optimized, the resulting model file is deployed directly to the vehicle to initiate the cold start phase.

Phase 2: The On-Vehicle "Cold Start"

Thanks to the homogenous architecture of our Vehicle-Cloud Data Foundation, the intelligent model is deployed seamlessly to the vehicle. Initially, it runs in "shadow mode"—operating silently in the background. It autonomously learns the user’s specific trip history and driving style. During this phase, it generates real-time thermal control strategies but does not execute them. Instead, these outputs and performance data are periodically uploaded to the cloud, providing the essential feedback for the next round of model iteration.

Phase 3: The On-Vehicle "Hot Start"

After several cycles of cloud-based optimization, the now-proven intelligent model is fully activated and "goes live." At this point, its recommended control strategies are actively executed by the vehicle. The live workflow operates as a continuous, intelligent loop: it predicts the conditions of the upcoming trip, calculates the anticipated heat generation from core components, formulates a tailored and proactive control strategy, and finally passes that strategy to the main thermal management system for execution. This enables a dynamic and intelligent response to the journey ahead.

To demonstrate the real-world impact, let’s look at the two functions with the most significant effect on driving range: battery heating and cooling. The following data comes from detailed measurements taken from production vehicles in our live customer projects.

In cold-weather environments, our intelligent thermal management system uses real-time driving conditions and trip prediction to proactively and gradually activate battery heating, avoiding aggressive and inefficient power draws.

• For short trips (≤10 km), where the effect is most pronounced, the system delivers energy savings of 5.7%.

• For trips between 10 and 20 km, the savings are 2.8%.

• Even on trips longer than 20 km, it still achieves a notable 1.2% in energy savings.

In hot-weather environments, the system applies the same intelligent logic, using trip characteristics to gently and efficiently activate battery cooling.

• Our test data shows energy savings of 6.6% on short trips.

• For medium and long-distance trips, the savings are 3.8% and 1.23%, respectively.

This real-world data provides clear validation of the EXD system's optimization capabilities across a wide range of temperatures and driving conditions. For the vehicle owner, these percentages translate directly into what matters most: a tangible increase in real-world driving range.

A Real-World Case Study: Intelligent Thermal Management in Action

Let's look at a typical driving scenario to see how this technology performs in a real-world environment.

▶ Scenario: Precise Battery Heating Control for a Short Winter Trip

On a cold winter day, a conventional thermal management system faces a common inefficiency during short trips. Because it has no awareness of the journey's end, it will often continue heating the battery right up until the car is turned off, wasting significant energy on "over-conditioning."

The EXD smart thermal management system solves this problem directly. By leveraging its predictions for the upcoming trip conditions, the user's driving style, and the vehicle's overall energy flow, it can precisely control the timing of the battery heater.

Crucially, the system intelligently anticipates the end of the journey and proactively shuts off the heater before the trip is over. This saves a significant amount of energy while still ensuring the battery remains within its optimal temperature range for the entire duration of the drive.

It is this intelligent and precise control that lies at the heart of the EXD system's ability to deliver significant, measurable energy savings.

Conclusion

The next wave of breakthroughs in EV efficiency and the overall ownership experience will be defined by intelligent, precise thermal management strategies.

The EXD intelligent thermal management system is engineered to solve these exact industry pain points, delivering higher energy efficiency, greater real-world range, and a superior user experience in all conditions.

And the value of our platform extends far beyond this single application. Our customers and partners are already leveraging the EXD Vehicle-Cloud Data Foundation to explore other high-impact scenarios, including intelligent diagnostics, smart climate control, and adaptive suspension.

We welcome more partners with deep domain expertise to join us in co-creating the next generation of data intelligence solutions and, together, build a future of smarter, data-driven mobility.