Creating vehicle components typically requires a significant amount of time and financial investment. Researchers at Graz University of Technology (TU Graz) have introduced a new approach that can reduce the development timeline for the powertrain of battery electric vehicles by several months. Under the guidance of Martin Hofstetter from the Institute of Automotive Engineering, a team is merging simulation models of components with evolutionary optimization algorithms. This AI-driven system automatically fine-tunes the entire powertrain—from power electronics to the electric motor and transmission—while adhering to the manufacturer’s technical specifications and considering crucial factors like production costs, energy efficiency, and spatial constraints within the vehicle. The OPED (Optimization of Electric Drives) software solution, which emerged from almost a decade of research at TU Graz, is already being successfully utilized by an Austrian automotive supplier.
The foundation for this automated optimization begins with entering the technical specifications that the powertrain must meet. These specifications cover aspects like power output, minimum service life, maximum achievable speed, and available space within the vehicle. “Electric drives comprise numerous components that can be designed in varied ways to meet the required specifications,” states Martin Hofstetter. “A minor modification to the electric motor, for instance, impacts the transmission and power electronics, making it highly complex to make optimal choices.” Further complicating matters is the lack of a singular perfect powertrain solution, as different manufacturers may prioritize factors such as production costs, weight, volume, or energy efficiency.
One day instead of several months
The OPED software significantly simplifies this complexity. It takes the technical requirements and simultaneously varies and combines around 50 design parameters, comparing the simulated powertrains against the manufacturers’ priorities. Unviable designs are eliminated while more promising options undergo further refinement. After hundreds of thousands of calculations and simulations, OPED identifies solutions that align closely with the manufacturers’ goals. They can then choose from a limited number of options to pursue for further development. “Tasks that engineers would normally take months to complete without AI support can be done in roughly a day with OPED,” explains Martin Hofstetter. “This enables development teams to concentrate on high-level decisions rather than spending their limited time on manual calculations and simulations.”
Optimization for an entire vehicle platform
The OPED system is also designed for flexibility. The researchers have incorporated CO2 emissions, both during the use of the powertrain and throughout its production in the supply chain, as optimization criteria. This approach ensures that sustainability is factored in from the earliest stages of development. Recently, Dominik Lechleitner has advanced OPED further by expanding it to optimize electric powertrains across entire vehicle platforms in his doctoral research. This method aids in identifying optimal components that can be utilized as common parts across different models within a platform, thus minimizing development and production costs.
“The OPED method is applicable to a broad spectrum of product development,” remarks Martin Hofstetter, “and we’re eager to collaborate with new industrial partners to tailor it to their specific challenges and objectives.”
For their innovation, Martin Hofstetter and Dominik Lechleitner received the VDI Prize 2024 from the Association of German Engineers. Additionally, Martin Hofstetter was honored with the Kamm-Jante Medal from the Scientific Society for Automotive and Engine Technology (WKM) and the 1st Prize of the Vehicle Association Anniversary Foundation, which is presented by the Automotive Industry Association of the Austrian Economic Chamber.
