The UK government’s Green Industrial Revolution will see a ban on the sale of new gasoline and diesel cars and vans by 2030—ten years earlier than previously planned—in a move which will impact all areas of the automotive industry. Alongside this, growing public concern about the environment is driving ever stricter regulatory action, meaning for automakers the green agenda will be front and centre in the 2020s. To achieve such ambitious targets, it will be critical to look beyond tailpipe emissions and address the problem from well-to-wheel.
Compressing 130 years of innovation into a decade
Modern cars are highly complex machines, developed iteratively over generations of models into a comfortable, convenient, safe and reliable mode of transport for billions of people. Innovations in fuel efficiency, emissions management and powertrain reliability have significantly reduced pollution levels but, to meet such ambitious targets, there is more work to be done.
To achieve such ambitious targets, it will be critical to look beyond tailpipe emissions and address the problem from well-to-wheel
Whilst the global electric vehicle (EV) market is on the rise, even the most optimistic estimates suggest global EV production will meet less than 25% of total demand in 2030, meaning in a decade’s time, a significant proportion of the cars on our roads will still have internal combustion engines (ICEs). So, to deliver on climate goals, the next decade needs to be focused on building and running every single engine out there—whether it’s an ICE, a plug-in hybrid or a pure EV—in the most energy-efficient way possible.
New breakthroughs in machine learning
This must start with rapidly speeding up vehicle development by at least an order of magnitude to accelerate time to market for more energy efficient engines. Many look to AI-based optimisation and calibration technologies to tackle complexity challenges and reduce development time. Today, new technology breakthroughs are finally putting a solution in reach.
To deliver on climate goals, the next decade needs to be focused on building and running every single engine out there—whether it’s an ICE, a plug-in hybrid or a pure EV—in the most energy-efficient way possible
New Innovations in AI technology have now become adept at quantifying uncertainty, managing high dimensional trade-offs and explaining outcomes using sparse and low volume data, making it well suited to addressing complex calibration and optimisation problems. Advances in ML technology that combine Bayesian Optimisation and Gaussian Processes to optimise calibration can be used in ways that are significantly faster and produce better outcomes than other techniques used today.
The benefits to automakers are numerous. Time saved in engine calibration will not only accelerate time to market for newly developed or recalibrated engines, but also result in cost savings whilst accelerating feedback on engine prototype design improvements. Furthermore, time savings mean domain experts can focus on other value-add tasks, such as devising new approaches to design which support environmental goals.
The simple truth is that automakers can’t afford to stand still: the consequence of missing emissions targets are multiple billions of Euros worth of fines and increasing consumer pressure. The road to 2030 will require innovative thinking but the technology is now emerging that can support OEMs as they accelerate towards a greener future.
The opinions expressed here are those of the author and do not necessarily reflect the positions of Automotive World Ltd.
Vishal Chatrath is Chief Executive and Co-Founder of Secondmind
The Automotive World Comment column is open to automotive industry decision makers and influencers. If you would like to contribute a Comment article, please contact email@example.com