Electric vehicle (EV) powertrain design is entering a new stage of maturity, where mechanical, electrical, and software systems must operate as one. Engineers are now expected to create modular architectures that meet diverse vehicle requirements while maintaining efficiency, reliability, and scalability.
Doug Cross, Board Advisor, Everrati
Thermal management, torque control, and system integration have become shared challenges that influence every stage of development. Software now plays an active role in balancing these demands, coordinating energy flow, and maintaining performance across changing conditions.
We spoke with Doug Cross, Board Advisor with Everrati, who brings more than 30 years of experience in automotive powertrain development, about the evolving relationship between hardware and software in EV design. Everrati specializes in redefining classic and high-performance vehicles with advanced electric powertrains, combining heritage design with modern zero-emission technology.
In this Q&A, Cross explains the complexities of modular architectures, how software supports thermal and torque management, and what it takes to validate systems that continue to evolve through updates.
Here’s what he had to say…
What are the key engineering challenges in designing modular EV powertrains?
The greatest challenge is the sheer variation that must be accommodated. Vehicles differ widely in available packaging space, with some able to take “chest-type” battery packs, while others only allow underfloor solutions. Motors may need to be configured longitudinally in a T-drive layout or transversely with parallel axis drives.
On top of that, gear ratios must span everything from providing strong low-end acceleration to achieving top speed.
Customer demands also vary. Desired acceleration defines how much torque is required at low speeds, which in turn influences whether an all-wheel drive or two-wheel drive solution is viable. Packaging becomes especially challenging in front-wheel drive layouts, where it’s necessary to work around the steering rack, anti-roll bar, and pedal box.
The engineering challenge is to cover all of these scenarios while still developing the minimum number of unique components. The more reuse you can engineer into the system, the more efficient the solution becomes, but that balance is difficult to achieve when every platform has its own quirks.
What thermal management considerations are essential when developing high-performance powertrains, and how does software offer support?
Every major component has its own thermal limits. Batteries often attract the most attention, but with their large thermal mass, they heat more slowly. Inverters, on the other hand, generate smaller losses but have much lower thermal inertia, making them particularly challenging to manage. Motors and inverters together demand robust cooling.
Modern EVs increasingly rely on multi-loop cooling circuits. These systems can be arranged in series, parallel, or hybrids of both. Tesla’s “Octovalve,” for instance, is believed to allow dozens of circuit permutations to balance inverter, motor, battery, and cabin loads simultaneously.
Software orchestrates this complexity. By deciding which cooling configuration to use at any given moment, it minimizes energy consumption and maximizes efficiency. The system can prioritize battery longevity in one situation and cabin comfort in another, always managing trade-offs in real time.
What are the engineering challenges in integrating the powertrain with other critical EV systems, such as the battery, inverter, and vehicle controls?
The EverDrive is an example of a modular electric powertrain, designed for flexible integration across multiple vehicle platforms, combines high-power density with efficient thermal management and scalable software control.
Integration begins with customer requirements. If an OEM specifies a 0 to 60 mph time of 4.5 seconds, we must simulate weight transfer, grip limits, and wheel torque to determine whether the target is achievable. Those system-level results cascade down to component-level requirements for the motor, inverter, and battery.
The challenge is making sure the control systems operate within achievable limits. If the vehicle controller calls for full wheel torque within one millisecond, but the hardware cannot physically deliver, you need robust error handling or better yet, realistic calibration from the outset.
This balance between control demand and hardware capability is where integration succeeds or fails. Over-ambitious software can push hardware into fault states, while under-utilization leaves performance on the table. Careful co-design of requirements, simulations, and calibration is essential.
In what way is software now responsible for optimizing core powertrain functions, such as torque delivery and energy efficiency?
Torque management is not new, as traction control systems have been shaping torque delivery for decades. What has changed is the precision and speed at which software can now act.
For example, in-wheel motors allow torque adjustments to be applied directly at the tire with millisecond response. That enables fine-grain control of traction, electronic stability control, regenerative braking, and even torque vectoring at a frequency impossible with traditional drivetrains that had to tolerate driveshaft wind-up.
Software is also one of the biggest key players when it comes down to efficiency. Regenerative braking in this case, which has been around since the first hybrids, remains a critical contributor by recovering energy that would otherwise be lost as heat.
Depending on the architecture, regenerative strategies can add anywhere from three to 15% of usable range. The challenge is to balance aggressiveness of regen with driver comfort, since extreme regeneration can feel unnatural. Software plays the decisive role in tuning that balance.
How is the shift from hardware-centric to software-defined systems changing the way engineers approach powertrain validation and testing?
Hardware and software have always co-existed in powertrains, and both require thorough validation. The difference now is that software is no longer static. In the past, a vehicle might run the same code from launch until the end of its life. With over-the-air updates, that model has changed completely.
Today, updates allow OEMs to add features, refine drivability, and mitigate durability issues during the vehicle’s lifecycle. But this creates new constraints. The latest software must remain compatible with an entire fleet of legacy vehicles.
This re-engineered GT40 has a software-defined electric powertrain that enables real-time optimization of performance, efficiency, and thermal control. (Image: SODA)
This means engineers now must consider whether older hardware can run newer code, whether the necessary sensors are present across all builds, and whether the sensor stack is consistent between generations.
Validation has expanded from being a one-time exercise at launch to an ongoing responsibility. Every update must be tested across multiple hardware and sensor configurations to ensure safe operation without undermining the customer experience.
What are the most critical considerations when scaling an EV powertrain solution from prototype to commercial production?
Prototypes are designed to prove potential. They only need to demonstrate that the concept works, and sometimes they do so intermittently with engineers stepping in to troubleshoot. Production systems cannot operate like that. They must work flawlessly under all foreseeable conditions.
Reliability and fault-tolerance are the dividing lines. A production system has to degrade gracefully. If a thermal system fails, the car should still limp home rather than shutting down completely. Delivering that requires exhaustive verification across all edge cases and environments.
The transition from prototype to production is not a simple scale-up. Rather, it’s an enormous effort in reliable engineering, fault detection, and fail-safe strategies. Typically, this is where the bulk of the development time is spent.
How do continuous software updates influence long-term powertrain performance and reliability, and what safeguards are needed to ensure safety and security?
Over-the-air (OTA) updates are a double-edged sword. On one hand, they allow OEMs to refine performance, add features, and address durability issues long after launch.
For example, a vehicle might launch with conservative torque limits until full durability testing is complete. Later, software can safely release the additional performance. Conversely, if an unexpected reliability issue arises, an OEM can temporarily limit performance until a service intervention replaces the part.
On the other hand, updates introduce risks. Each new software build must be validated across every affected vehicle variant to ensure no unintended interactions. Cybersecurity is paramount. Bad actors targeting OTA updates could have catastrophic consequences, so compliance with rigorous security standards is non-negotiable.
Ultimately, continuous updates extend the life and value of EVs, but only if the safeguards such as validation, compatibility checks, and robust cybersecurity are built into the engineering process from the beginning.
Filed Under: FAQs, Featured Contributions, Q&As, Thermal Management