Interview with Martin Baumann, BMW
“Ensuring data trustworthiness involves three key aspects”
Martin Baumann has been with BMW for eight years and is now a Hardware Development Engineer working on zonal power distribution units, with deep expertise in the virtual development of low-voltage power systems, system identification, impedance spectroscopy, hardware prototyping and validation in BMW iX-based labcar environments.
Martin Baumann
As vehicle architectures grow more complex and energy-hungry, power distribution is becoming a strategic challenge. Martin Baumann, Development Engineer at BMW, explains how data-driven methods are changing electronic power distributor design.
As vehicle architectures become more
software-defined, function-dense and electrified, the demands placed on
low-voltage power systems continue to rise. That is making the development of
electronic power distributors far more data-driven than in the past.
Martin Baumann, Development Engineer at BMW, works on zonal
power distribution units and focuses on robust, efficient electrical systems
for next-generation vehicles. In his keynote at Bordnetzkongress
2026, he discusses how measurement data, simulation and validation can
improve design quality, speed up development and support better engineering
decisions across platforms and variants.
Ahead of the event in Ludwigsburg, we spoke with him about
the challenges, opportunities and current limits of data-driven development in
electronic power distribution.
ADT: Looking ahead five years, what will be the single biggest challenge for the wiring harness and EDS industry,
and why?
Baumann: From my area of expertise in electronic power
distribution, the biggest challenge will be managing increasing vehicle
requirements in terms of energy demand and functionality. It will also involve
incorporating key learnings from products based on the previous architecture
and introducing the technical changes associated with future electronic fuse
products entering the market. As new features are introduced, it is critical to
thoroughly analyze their impact on vehicle requirements. Furthermore, the
insights gained from the current implementation of electronic fuses in Neue Klasse vehicles will significantly influence the
design and development of the next generation of vehicles.
What is the one capability companies must build now to
make data-driven harness development scalable?
From an OEM perspective, it is essential to ensure the
high-quality acquisition of measurement data from series production vehicles,
which must be securely stored in cloud infrastructures. This data then requires
efficient computational processing to translate those insights into concrete
design requirements, such as wire sizing, connector specifications, and fuse
characteristics. For developers, precise characterization of ECU parameters,
particularly thermal behavior, is necessary to guarantee the integrity and
correct processing of the transferred data. Semiconductor manufacturers must
provide components with highly accurate measurement and diagnostic capabilities
to support this ecosystem.
Which data sources actually move the needle for power
distribution design, such as field data, simulation, end-of-line tests, or
fleet diagnostics?
Fleet data is particularly valuable, including measurements
of voltages, currents, and temperatures. The current drawn by individual fuses
offers insights into load characteristics and stresses on power distribution
components, such as wiring. The measured data can also be used to simulate
system behavior under rare overlapping failure scenarios. Additionally, fleet
diagnostics enable targeted fault localization and rapid troubleshooting, while
predictive models can be developed even before faults occur, enhancing
proactive maintenance strategies.
How do you ensure the data is trustworthy enough for
engineering decisions, especially across variants and platforms?
Ensuring data trustworthiness involves three key aspects:
concept development, validation, and testing against fleet data. During concept
development, measurement approaches are designed to meet strict accuracy
requirements. Validation takes place at multiple levels – simulation,
component, and system – and includes sensitivity analyses with regard to
environmental influences such as electromagnetic interference, temperature
variations, aging effects, production tolerances, and humidity. This comprehensive
approach ensures the reliability of the data used for engineering decisions
across different vehicle variants and platforms.
Can you share one concrete example where data-driven
development changed a design choice, such as architecture, sizing, or
protection strategy, and what the outcome was?
A concrete example is the design of overcurrent protection
mechanisms for an electronic fuse used in high-current applications. Extensive measurements at BMW test facilities
approached worst-case simulation scenarios, allowing the maximum current
profile to be integrated into the design of the fuse’s protection mechanisms.
Based on this current profile, adjustments were made to power semiconductor
selection and to hardware- and software-based overcurrent trip-curve parameters.
The choice of power semiconductors also affected the thermal protection
strategy, demonstrating the holistic influence of data-driven insights on
design decisions.
Where are the limits today, and what still cannot be decided
with data and still needs expert judgment or physical testing?
Rare events, such as vehicle crashes, inherently involve
uncertainty due to their low frequency of occurrence and the impracticality of
extensive testing. For these exceptional failure cases, expert judgment remains
indispensable, alongside physical testing, to complement the limitations of
data-driven methodologies.
Finally, what do you personally hope to take away from
the Bordnetzkongress 2026 in Ludwigsburg?
I aim to deepen my understanding of industry trends in
electronic power distribution – especially upcoming developments, key
challenges and ways to improve collaboration throughout the technical
development process. I also hope to identify data gaps from an OEM perspective
that, if addressed, could support better design and faster development cycles
while maintaining the same high level of quality.
