The $40 billion blind spot: NSF CAREER Award corrects the broadband map that leaves rural communities stranded

Shaddi Hasan, an assistant professor of computer science, has been working on a question that sounds simple but has frustrated policymakers for years: Who in the United States actually has reliable internet, and who doesn’t?

His recently awarded National Science Foundation Faculty Early Career Development Program (CAREER) award will advance that work for the good of rural and indigenous communities across the country.

What disconnection costs is not abstract. Reliable internet access now underpins telehealth, telework, and economic development in communities that need it. Without it, learning suffers as well; the pandemic exposed a homework gap when students could not get online at home.

“This isn't research happening at a safe distance from the problem. These are real communities with real stakes, and Shaddi's work gives them tools to fight for what they're owed,” said Christine Julien, head of the Department of Computer Science. “The CAREER award is the most prestigious recognition NSF gives early-career faculty, and it's hard to imagine a better example of what it's meant to reward.”

A map to nowhere for some

Hasan’s project takes aim at a problem hiding in plain sight. The Federal Communications Commission maintains a national broadband map built from data that Internet service providers submit twice a year. On screen, the result is reassuring.

"If you look at the map, the whole country has the Internet, the problem is solved, and we can all go home," Hasan said. "But those maps decide where funding goes, and the companies drawing them are the same ones competing for that money."

The stakes are considerable. The federal $42.45 billion federal Broadband Equity, Access, and Deployment Program is meant to expand high-speed, reliable internet infrastructure to unserved and underserved communities. Yet Hasan notes that providers awarded contracts to build the infrastructure often cover only 80 or 90 percent of a target area, leaving out the same rural and Indigenous communities that have been left out for decades.

The program creates an incentive for providers to report service in places they do not fully reach. That can make a community look connected when it’s not and disqualify it for funding to connect through other means. In Hasan's early work, his team found locations listed as covered on the broadband map that on the ground had no service.

The result: Residents of many rural communities across the U.S. can't get crucial internet services, and localities without broadband can't recruit businesses that would bring more jobs to the community. 

Starlink satellite service, which touts itself as “home internet in rural and remote areas,” has not closed the gap. A household can place a dish almost anywhere and connect, but the low-Earth-orbit network of satellites can’t carry every disconnected American at once, and the monthly cost puts the option out of reach for most of the low-income households that need it. 

Hasan describes satellite service as serving "anyone anywhere, but not everyone everywhere,” the title of a paper his lab recently presented on the subject at the ACM Workshop on Hot Topics in Networks. 

Better data, better decisions

Providers do not build broadband infrastructure at random; they follow the economics of population density, existing roads, and nearby infrastructure. By combining network measurements with remote-sensing observations and machine learning, Hasan aims to reverse-engineer those network planning business models and predict where the maps are most likely wrong. This would aim scarce auditing resources where they can do the most good.

The payoff of this approach, Hasan says, is better data for better decisions. Correcting the official national broadband map is arduous. To file challenges, residents must route evidence, including photographs of missing infrastructure, through a government agency or an approved nonprofit. But the data Hassan’s team will gather will equip communities to challenge existing maps, make the case for broadband investment, and hold neglectful companies to account.

"People truly do rely on these systems for a good quality of life," Hasan said.

To fulfill the CAREER award's educational mission, Hasan plans to scale up a partnership with the Nez Perce Tribe in north-central Idaho and other rural and Indigenous communities, helping them prepare the technical filings that federal broadband programs demand.

Working alongside computer science professor of practice Margaret Ellis, Hasan’s undergraduate students have already built open-source tools that one tribe now uses to document its internet coverage. This is one example of a classroom assignment that produced infrastructure a community can depend on.

That is the through line of Hasan’s work: Not connectivity handed down from a distance, but data and tools that let communities advocate for themselves.