FCC working group proposes using AI for better broadband maps

Solving the digital divide in the US has gained a lot of attention in the past year thanks to the COVID-19 pandemic, which forced much of the country to work and study from home as a way to contain the spread of the virus. But the pandemic also underscored the need for accurate broadband data maps because, without accurate data, it's nearly impossible to determine where the country lacks competitive services.

The FCC's Technological Advisory Council Working Group on Artificial Intelligence (WGAI) may provide some solutions to this dilemma. A white paper the group submitted in January details how the WGAI believes AI can help the FCC in a number of ways, and included specifics on how AI could help close the digital divide. The group called AI a "game changer" for the industry and recommended the FCC create a task force on AI and also conduct a request for information (RFI) in this area to determine more specific examples of how AI can be used for broadband mapping.

Lisa Guess, head of global presales engineering at Ericsson's Cradlepoint, was one of the chairs in charge of the FCC's AIWG. Guess said that the group looked at how AI could improve FCC processes as well as where federal investments should be made in AI. "Part of our strategy was to innovate and unlock opportunities to closing the digital divide," Guess said. "We recommended the FCC cast a wide net."

To be clear, the outcry for new broadband coverage maps started well before the pandemic. The FCC's existing maps are primarily based on data submitted by providers themselves, and many have argued they are not accurate. Indeed, an FCC report released in 2019 found that Verizon, T-Mobile and U.S. Cellular provided inaccurate coverage maps about 40% of the time.

But the topic has gained steam in recent months following Congressional mandates – and funding – for the FCC to improve its maps.

Adam Drobot, chairman of the board of OpenTechWorks and also a chair of the FCC's AWIG group, said that AI could provide the commission with analysis tools and actionable data. "This is about pulling together data and putting it together in an intelligent way with query tools that allow you to understand the data," he said.

One of the benefits of AI is that it can be used to look at coverage data in different ways. For example, it could look at different geographies or population densities or economic factors. That type of criteria could then be used to help determine what type of broadband technology makes the most economic sense in that area. "It's got to be affordable," Drobot explained. "If a certain population has a high poverty level, then it is important to understand that and bring the right tools."

So what's next? The FCC will study the contents of the paper and decide whether to move forward with any of the recommendations, including using AI to create better broadband maps.

Guess said she believes that the FCC will likely agree to some of the recommendations in the paper.

The recent appointment of Jessica Rosenworcel as acting chairwoman of the FCC may help the AWIG's white paper get more traction, particularly when it comes to creating better broadband maps. "My feeling is it will get a reasonable reception from her," Drobot said.

Rosenworcel recently formed a task force to review the mapping problem, but officials involved in the effort said they don't expect to obtain updated maps until next year.

— Sue Marek, special to Light Reading. Follow her @suemarek.