Key takeaways from Powering the future: Power Electronics

R&D priorities

A first set of takeaways relates to the technological advances needed to further improve efficiency and performance.

1. Wider deployment of wide-bandgap semiconductors
A broader application of wide-bandgap semiconductors can enable smaller and more efficient devices with higher power densities. Technologies such as bi-directional GaN devices and SiC inverters were highlighted as important developments that can significantly improve power conversion efficiency.

2. High-frequency magnetics
High-frequency magnetics are another key area of innovation. Progress in this domain makes it possible to achieve very high efficiencies even at lower power levels, which is essential for many emerging applications.

3. AI-powered design and R&D tools
Artificial intelligence is increasingly finding its way into engineering workflows. AI-powered tools, including design automation and advanced machine learning, can support smarter R&D processes and help engineers navigate complex design challenges more efficiently.

Finding the sweet spot between optimisation and manufacturability

A recurring theme during the event was that the best electrical design is not always the best real-world solution.

Converters must not only perform well in simulations but also be manufacturable, qualifiable, serviceable and scalable. As a result, identifying the right balance between electrical optimisation and manufacturability is a central challenge.

This balance needs to be considered from the earliest design stages.

Industrialisation: priorities for scaling innovation

Bringing power electronics innovations to market requires addressing several practical challenges linked to large-scale deployment:

• Hybrid cooling solutions
• Smart adaptive control algorithms
• Reliable and safe packaging
• Design for serviceability

Reliability starts with better design and collaboration

Another important message from the panel discussion was the need to avoid inventing solutions in isolation. Collaboration across the ecosystem is key to accelerating progress.

Participants also stressed that middleware alone is not the differentiator. Instead, improving reliability requires stronger design and simulation efforts at the start of the development process.

For example, achieving accurate lifetime predictions requires designing for junction temperature variation, not only for junction temperature itself.

AI factories: a moonshot for scaling power electronics

Looking further ahead, the concept of “AI Factories” was presented as a potential moonshot to accelerate the scaling and industrialisation of power electronics innovations.

Such initiatives could support the development of:

• Flexible and scalable power systems
• Open software layers with plug-in middleware for LVDC systems
• Architectures based on bi-directionality and reconfigurable topologies
• Automated maintenance

A key priority is strengthening electrification competences and skills across the entire value chain — from researchers and product designers to installers and service professionals.

In addition, digital twins can help simulate and validate new applications quickly and reliably, accelerating the development and deployment of new solutions.

Looking back on the event

With more than 80 participants, the event provided a valuable opportunity to exchange insights and connect with others active in the power electronics community.

We hope the discussions and perspectives shared during the event will contribute to well-considered strategic and policy choices in the years ahead.