ISSUE 024 March/April 2024 Frauscher x Porsche 850 Fantom Air dossier l In conversation: Michael Fischer l Polymers focus l Potting & encapsulation insight l BEDEO van conversion l Power semiconductors insight l Inverters focus

15 Michael Fischer | In conversation E-Mobility Engineering | March/April 2024 This led to a range of research directions, including particle filters for gasoline engines, or new, selective catalytic reduction (SCR) and lean NOx catalyst (LNC) solutions, aimed at NOx aftertreatments for both gasoline and diesel engines. AI for cleaner powertrains Before long, Fischer’s scope expanded to include new forms of software-defined solutions for directly reducing emissions during the combustion process. “For instance, I soon took on some projects evaluating the use of artificial neural networks for engine control, as well as aftertreatment control. Some key objectives were to intelligently estimate NOx emissions being produced by engines, and to control the exhaust gas-recirculation rate and NOx aftertreatment systems to reduce them most effectively,” he recounts. “Those projects have been highly scientific, and that’s why I decided to do my PhD thesis on neural networks for transient NOx estimation, all outside of work hours, but using project results from my research at Honda.” In that PhD, Fischer used a selforganising map – a type of artificial neural network (ANN) trained using unsupervised learning – to produce virtual sensors capable of closely estimating transient NOx emissions, based on sensor signals commonly available in mass-production engines (specifically engine speed, injected fuel mass, Lambda, mass air flow, boost pressure and exhaust-gas temperature). Estimations were performed using local linear functions, and their quality showed they could replace an engine-out NOx sensor, making for potentially significant cost advantages to powertrains incorporating such technology into their control software. “We developed advanced control strategies, and elaborated on how those could be applied in hybrid cars, especially in optimising fuel consumption as engines could be run Honda was already producing HEV cars and powertrains when Fischer started working at Honda R&D Europe in 2006 independently of traction requirements – for me, that was the big leap into e-mobility. Honda already had great experience in hybrid powertrains and then it accelerated for everyone following the ‘Dieselgate’ emissions scandal a couple of years later.” Circularity with three points Through such experiences, Fischer now steers Honda R&D Europe’s Energy & Automobile division with a three-pronged approach to sustainability targets, combining mechanical engineering, materials science and digitalisation. “Digitalisation in particular is a huge buzzword, but we really do try and keep track of how software can help us make mobility a more sustainable world,” he says. “Take, for example, the target of resource circulation. One primarily views this as a material challenge: what kinds of materials are you using in, say, batteries, where are you extracting them from, how can you avoid materials that are environmentally harmful or difficult to harvest? “But there’s also a strong digital component: through software databases, we can track each of those parameters at each stage of a product’s lifecycle. We can track and verify that the companies in charge of material extraction are doing so through sustainable and humane practices.” Additionally, Honda R&D aims to make sure such data cannot be defrauded or cheated, to ensure components that are developed for production are genuinely sustainable from life to end-of-life, and not mere greenwashing. This has spurred considerable focus by Fischer and his team into blockchain. As blockchain-powered ledgers cannot be modified, and are inherently cyber secure, blockchain or similar approaches show immense potential for digital solutions on the traceability of goods and energy. EVs as home devices As EVs become increasingly softwaredefined, there are growing suggestions that cars will one day be used as devices in a similar way to smartphones. Fischer cautions, however, against certain impracticalities of such statements. “An excess of infotainment and personal information devices inside EV cabins can be a huge and unsafe distraction for drivers, certainly as long as we’re the ones doing the driving and not some self-driving autopilot system,” he says. However, there is growing interest in many markets that there should be ways of using EVs and their great energy storing and charging capabilities to advantage in both mobility and home applications, particularly to minimise the amount of energy taken from the grid. To that end, Honda places a lot of focus on energy and charging intelligence, which Fischer points to as being one of the most undervalued and underrepresented topics in discussions about e-mobility engineering.

RkJQdWJsaXNoZXIy MjI2Mzk4