Keynote Speakers

Reliability is a key factor when designing automotive power electronics, and extensive thermal modeling activities are involved to verify if a proposed design can withstand the harsh environment, combined with the loads expected. Often one of the cost drivers is the cooling subsystem, and the quality of thermal models directly impact how much it needs to be overdesigned. Part of the available design space will be lost to the guard band covering both the uncertainty of temperature predictions, and manufacturing tolerances. To improve accuracy of thermal models, dedicated model verification activities or model calibration can be included in the engineering workflow. To ensure all produced parts are according to specification some products employ 100% thermal performance testing in our factories.

My hobby is electronics, I love playing with electronic instruments, trying to measure the unmeasurable. About 10 years ago I got captivated by the world of thermal measurements and been serving my fellow reliability colleagues (and others) ever since. In my talk I will recall my encounters with topics like (thermal) model accuracy, model vs. measurement vs. reality, suitability of a model to a purpose, all in the context of automotive power electronics. I might even add a quick glimpse of the bright future…

Driven by the ever-increased societal needs for digitalization and intelligence, such as autonomous driving, “Smart-X”, “AI in all”, the demands for mission critical electronics components and systems are growing fast. To realize “performance and lifetime on demand” of these mission critical electronics, Digital Twin (DT) will play an essential role. DT consists of mainly 3 parts: physical products in real space, digital products in virtual space, and connectivity of data and information that tie the virtual with real product together. This talk will focus on the major success factors of developing a reliable virtual twin via multi-scale and multi-physics simulation and optimization, covering the whole product creation, qualification and application conditions. Some challenges for future developments will also be highlighted.

The performance of light-emitting diodes has significantly improved in the last three decades. The luminous efficacy of the first white phosphor LED constructed in 1996 was comparable with an incandescent bulb, whereas today, the best-in-class commercial LEDs exceed 230 lm/W approaching a practical limit. The primary driver of the technology shift from conventional to solid-state lighting has been the need for high-efficiency light sources to save energy costs. The steadily increasing LED efficacy and decreasing luminaire prices enabled lighting upgrade projects with relatively low return-on-investment. As the baseline for efficiency improvements increases, the potential energy cost saving decreases in absolute value, rendering energy upgrades less attractive in the future. Beyond energy saving, LED lighting can deliver additional benefits to customers.  Superior light quality or emission spectrum tailored for physiological responses may be more valuable to the market than a further 10-20% efficacy improvement. In this talk, general and horticultural lighting use cases will be presented to highlight the tradeoffs between efficacy and light quality and provide insights into LED lighting trends.