Most plug-in hybrid cars have just several basic driving modes designed for general driving scenarios on the highway or in city traffic. A new system can actually learn from driving trips to balance the use of electricity and fuel in the most fuel-efficient manner.
This futuristic vision of hybrid cars adapting to become more fuel efficient comes from engineers at the University of California, Riverside. Their special hybrid energy management system uses machine learning software to improve vehicle fuel efficiency based on road and traffic conditions. By comparison, basic “binary” hybrid car modes make the car start driving in all-electric mode and continue until the battery has been depleted. Then the gasoline engine takes over. Driving tests on a 32-kilometer commute in Southern California showed that the new “learning” system achieved average fuel savings of almost 12 percent compared with today’s binary-mode systems.
“Our current findings have shown how individual vehicles can learn from their historical driving behavior to operate in an energy efficient manner,” said Xuewei Qi, a Ph.D. candidate in electrical and computer engineering at the University of California, Riverside, in a press release. “The next step is to extend the proposed mode to a cloud-based vehicle network where vehicles not only learn from themselves but also each other.”
The idea of networked cars wirelessly sharing information that makes them more fuel efficient could become a reality in the not-so-distant future. But for now, the university has filed patents based on the research published in the 5 February online edition of the journal Transportation Research Record.
Perhaps the biggest caveat for this study’s findings comes from how the researchers compared their special learning mode with the binary hybrid car modes. Many commercial hybrid cars actually offer several driving modes that don’t necessarily just run the electric motor until it’s depleted. For example, the 2015 Honda Accord Hybrid uses an all-electric mode only when starting from a stop or braking. Otherwise it runs on a combination of both electricity and gas—or the engine by itself once it reaches highway cruising speeds.
In that sense, it’s probably worth taking the reported fuel efficiency improvements in the study with a pinch of salt. But the general idea of having an energy management system capable of learning from historical driving trip data certainly seems promising. That concept could truly become revolutionary if the vision of smart cars talking with each other becomes widespread.
Jeremy Hsu has been working as a science and technology journalist in New York City since 2008. He has written on subjects as diverse as supercomputing and wearable electronics for IEEE Spectrum. When he’s not trying to wrap his head around the latest quantum computing news for Spectrum, he also contributes to a variety of publications such as Scientific American, Discover, Popular Science, and others. He is a graduate of New York University’s Science, Health & Environmental Reporting Program.