On Thursday, the U.S. Senate approved an immigration reform bill by a 68–32 vote. The bill faces an uncertain fate in the House, but polls show steady public support for major immigration reform. So it’s worth taking a look at one of the least-discussed provisions of the current Senate bill—one that could significantly affect the hundreds of thousands of people who travel by air every day in the U.S.

The part of the bill in question deals with how travelers enter and exit the U.S. at major airports. It would require the State Department to set up a “mandatory biometric exit data system” within two years at the ten U.S. airports with the most international passengers. Within six years, the system would be extended to the rest of the FAA’s “Core 30” airports.

In plain English, this means that everyone traveling on an international flight from the U.S. would go through passport control when they depart as well as when they arrive. Travelers who aren’t U.S. citizens would have their photograph and fingerprints taken.

Supporters of “biometric exit” say it will allow the federal government to more quickly identify foreigners who have overstayed their visas, a group of people that by some counts comprises 40 percent of all illegal immigrants in the U.S. The practice is already standard in most of the world—look at your passport, and you’ll see that almost every trip has both an entry stamp and an exit stamp.

A useful system, it would seem, except for one unavoidable problem: our airports aren’t designed to support it.

Unlike most of western Europe, for example, where certain terminals feature immigration checkpoints immediately before security, it’s not uncommon in the U.S. for a flight to Tampa to board alongside a flight to Tokyo. International passengers in America usually check in at the same desks, wait on the same security lines, and in many cases depart from the same terminals as domestic ones.

It’s a reality that stems from certain fundamental aspects of the aviation industry unique to the United States.

The predominant business model at American airports, where the airlines themselves own or lease most of the infrastructure, means that air carriers have tremendous leeway to design terminals however they want. This is especially the case at “fortress hubs” where one airline controls most of the airport operations.

Add to that the prevalence of the “hub-and-spoke” system, and the result is airports that quickly funnel passengers from small commuter planes to long-haul international flights in a matter of minutes.

The busiest American airport for international travel—New York’s Kennedy International—is a major exception to this. But plenty of airports that will be required to implement a visa exit system, from Dulles outside Washington, D.C., to Atlanta’s Hartsfield-Jackson International, are designed this way.

As a result, fundamental questions about how the exit system would work remain unanswered.

The most pressing issue is that it isn’t clear where the checkpoints could even be constructed.

Putting them before security would require a significant reconfiguration of how terminals are laid out, which airlines are likely to oppose. And that option wouldn’t cover the travelers who change planes from domestic flights to international ones without leaving the terminal.

A passport check at the gate, meanwhile, while not requiring any major reconfiguration, would delay boarding times significantly.

There are ways to fix these issues. One solution would be switching to a “common-use” model, where more of the airport infrastructure is owned by the airport itself and shared by the airlines.

Such setups are popular in Europe, where most flights are transnational, because it makes it easy to organize airplanes by destination and not by airline. Applied to the U.S., it would give airports the flexibility needed to install the biometric checkpoints. Some airports here, most notably McCarran International in Las Vegas, have already adopted the system.

But with U.S. airlines so heavily invested in their chosen hubs, it’s unlikely the whole industry would simply switch over.

A more feasible solution is a “biographic” exit system that would record the information already found on foreigners’ passports. It wouldn’t be quite as accurate as the biometric system, but neither would it require such a substantial infrastructure investment. The Obama administration has already completed successful test runs of such a system at four crossings along the Canadian border.

Whether such a compromise would be enough to satisfy the border-security hawks in Congress remains unclear. With many Republicans already skeptical of the bill, it’s unlikely the language will be dropped as debate begins on the House side.

Barring a sea-change in the way our airports operate, however, there’s no biometric solution that won’t result in longer lines and even more frustrated passengers.