Researchers at Carnegie Mellon University have invented the first software that predicts the battery charge of electric vehicles in the different climatic and geographical conditions that impact their energy use, determining the best place for charging stations — including sites on the ground for terrestrial vehicles or vertiports for electric jets and other aircraft to take off and land.
The new model looks at wind, elevation and other environmental factors to generate more holistic predictions of the energy needs of electric cars and aircraft. The software could potentially enhance the charging grid by providing dynamic data about the best locations to place stations. The patent application for this software, INCEPTS, and for an associated method was published by the U.S. Patent and Trademark Office on April 15.
A recent study showed that one in five Californians who bought an electric vehicle later returned it, mostly due to the hassle of charging their cars on the road. The co-inventor of this innovation, Matthew Guttenberg, described a similar issue that, somewhat ironically, almost derailed a trip to a recent summit on electric vehicles: "A group took a ride from Stanford down to where the conference was in Los Angeles, and they nearly found themselves calling roadside assistance because they almost ran out of charge," he told The Academic Times. "The onboard estimator wasn't properly telling them the range."
"The demand for electric vehicle charging infrastructure is going to be paramount," said Seth Boudreaux, associate director of CMU's Center for Technology Transfer and Enterprise Creation, in an interview with The Academic Times. In the U.S., demand for electric vehicle charging energy is predicted to increase from under 5 terawatt-hours in 2020 to about 107 terawatt-hours in 2035. "Most of the technologies we see, they're cutting edge. But in this case, the market is defined and large, the customers are many and diverse, and the readiness level is already quite high … In a nutshell, the team has a really unique technical approach in a very hot market."
"One issue that persists with existing methods is that they treat the battery as a black box," the co-inventors wrote in the patent application. Guttenberg said that current models may overestimate the battery range of a vehicle by up to 20%, much like how gas mileage can decrease by 10%-40% in stop-and-go traffic.
"The way chargers are being placed right now is pretty much ad-hoc," said co-inventor Venkatasubramanian Viswanathan, an associate professor at CMU, in an interview with The Academic Times. "There's no grid system or formalized way to do it, for the most part.
"Volkswagen had to invest over a few billion dollars in putting up electric charging infrastructure across the U.S. as part of the Dieselgate settlement [in 2016]. We talked to state senators and utility companies to plan … but there was not enough momentum around electrification" to make a comprehensive plan for the future, Viswanathan added. Now he thinks INCEPTS could help the people who are developing the electric vehicle infrastructure of tomorrow determine the best real-world placement of chargers.
INCEPTS looks at the battery in a dynamic way by considering local weather and traffic conditions that inherently affect the charge of an electric vehicle. The co-inventors describe it as a plug-in model: a user of their invention, such as an electric car company or a city planner, inputs the route of a vehicle into INCEPTS, and the software analyzes environmental factors to decide how much battery power the vehicle needs in different types of terrain.
INCEPTS can also simulate the battery charge required by entire fleets of electric vehicles. Thanks to the power of high-performance computing, the software program can easily estimate the battery charge of thousands of individual vehicles in widely divergent settings. By accounting for wind speed or mountainous terrain, the software could in turn inform a grid, allowing it to better meet the high energy demands of electric vehicles.
As an example, an electric pickup truck was capable of driving about 15% further than the control after INCEPTS analyzed its route. Elevation alone led to a difference of 60 watt-hours per mile, which translates to a vehicle traveling 12 fewer miles than the control per charge.
The co-inventors note that the degree of difference INCEPTS calculated between routes at extreme ends of the spectrum was surprising, though somewhat expected. For instance, a vehicle that lost nearly 2,300 feet of elevation in one trip could travel 30% further than one that gained the same amount of elevation in a single charge.
"Extreme cases are very important … because you're not designing for the average case," Viswanathan noted. He added that consumers most often look at the extreme limits of a car when shopping for an electric vehicle, meaning that variability matters. So do incentives, noted the innovation's third co-inventor, CMU researcher Shashank Sripad. Government rebates can help to alleviate the range anxiety a new electric vehicle owner may feel. Sripad points to different rates of ownership even in Nordic countries, where electric vehicles are already relatively widespread by global standards: Electric vehicles had a 54% market share in Norway in 2020, where the government offers big subsidies for new owners, compared to only 16% in Finland, where rebate programs are less robust.
Tools like INCEPTS can help city mayors and planners understand the constraints of electric vehicles. "You want [the technology] to first be economically viable, or else it won't survive. But ultimately, you want electrification to penetrate to all communities, not just to some," Viswanathan said. Equitable access to chargers is important, as building developers or other stakeholders may not be willing to place electrical charging stations in all neighborhoods of a city.
After charging grids are established for electric cars and other types of ground transport, the next big step is to look toward the sky at vertiports. The co-inventors noted that a few select areas around the world are pioneering this technology, including Los Angeles in the United States and Melbourne in Australia.
Someday, INCEPTS may be able to provide a map for placing vertiports above Earth. "Urban air mobility is a much more critical problem," said Viswanathan. "This patent is uniquely positioned to address that market in ways that basically nothing else today can do."
"One of the things we try to highlight is that the software is built to be very modular to anything. We don't want it to just be something that we can only use for terrestrial vehicles or electric aircraft," Guttenberg added.
The application for this patent, "Method for determining optimal placement of electric vehicle chargers," was filed Oct. 9, 2020 with the U.S. Patent and Trademark Office. It was published April 15 with the application number 17/066755. The earliest priority date was Oct. 9, 2019. The inventors of the pending patent are Venkatasubramanian Viswanathan, Matthew Guttenberg and Shashank Sripad, Carnegie Mellon University.
Parola Analytics provided technical research for this story.
Correction: A previous version of this story misidentified the patent application as the patent. The error has been corrected.