All of the major autonomous vehicle companies–your Waymos and your Cruises and your Argos and your Ubers and your Voyages and your Pony.ais and a whole bunch of other companies you’ve probably never heard of—test in urban areas, interstates, and suburban-style roads. Few, if any, dedicate any time whatsoever to testing on rural roads.

It might be tempting to think this isn’t a big deal because if a human driver can master the windy, chaotic streets of San Francisco then surely a simple two-lane country road wouldn’t be a problem. But, the reality is more complicated than that. For one, many AV systems use high-definition maps to supplement sensor readings from the car itself to pinpoint their exact location, and rural roads have not been mapped in HD. Plus, rural roads tend to be poorly maintained. A lot of them have faded or non-existent road markings, which present serious challenges to computers trying to make sense of the world around them.

That is why Texas A&M is to develop a self-driving algorithm that can handle the unique challenges rural roads present. The university’s newspaper, The Battalion, has more:

Rural roads present a different challenge for autonomous vehicles. In urban areas, high definition maps are readily available. These are very detailed, are updated daily if not hourly, and are precise enough to enable the navigation of autonomous vehicles. However, these maps, along with clear signs and road markings, are not available for many rural roads. Additionally, environmental factors have different impact on rural roads, said Reza Langari, department head, professor and holder of the J.R. Thompson Chair in the Department of Engineering Technology and Industrial Distribution.

“Rural roads are subjected to environmental degradations and variations that interstate highways and arterial roads are not,” Langari said. “They are also designed and maintained differently … Autonomous vehicle technology needs to be more robust. In rural areas road boundaries are less defined, and this is more challenging, having to add capabilities.”

Doing this work, though, is expensive. The researchers estimate the car alone outfitted with all the necessary technology will cost $300,000 at the very least, “plus significant costs for development, manpower and data collection and analysis,” according to The Battalion. But the university, in conjunction with researchers from George Washington University and the University of California-Davis, received a $7 million grant from the Department of Transportation to do the study.

In addition to testing on rural Texas roads, the study will also test on urban and suburban roads in Washington, D.C. and Virginia to see how the algorithm performs under different conditions.

Obviously, private companies investing hundreds of millions if not billions of dollars into AV research want to make their products available in the most populous areas so they can make the most money. But, AVs could have just as much, if not more, benefits in rural areas, albeit to fewer people, by freeing up driving time—of which there is a lot—for other tasks. So it’s good to see federal funds researching something the private market isn’t yet supporting.

But we won’t be getting rural AVs any time soon. The study itself is slated to last for four years:

The project is scheduled to run for four years, and the year will be spent working on vehicle development, which will be done at A&M with input from all team members. The second and third years will be spent doing testing and data collection. Talebpour plans for the cars to run seven days a week and eight to ten hours a day, providing vast quantities of data to be analyzed. During the fourth and final year, the data will be examined to see what went well and what went poorly, with tweaks being made and those improvements retested.

We’ll check back in around 2024 or so to see how it went.