‘Damn, It’s That Cow Again.’ — The Border Patrol and Technology

Drones, tunnels, and mechanical dogs

By David North on February 10, 2022

A robot dog showing off capabilities during testing in Lorton, Va. Photo by DHS Science and Technology Directorate.


Some decades ago I was doing some research at the U.S.-Mexico Border. It was nightfall and I was at some sector headquarters watching the Border Patrol agents watch their then fairly new systems of sensors that were designed to help them detect illicit northward movements of human beings.

“Damn, it’s that cow again,” one of the agents said as he watched the readouts for one particular section of the border.

“How can you tell?” I asked.

He explained that an incoming illegal alien would set off an alarm right at the border, then another maybe 100 yards further north, and then another one, all on a logical path into the interior, and all at a timing that would suggest a person walking.

He was looking at the middle of a set of three alarms, and someone, or rather something, kept setting it off without the tell-tale similar signal from the next sensor. He suspected, based on past experience, that it was a grazing cow. I am sure he was right.

He was also using a low-tech way of managing a high-tech piece of equipment; the sophisticated sensor system could not tell the difference between a cow and a human, but an agent could.

It was my introduction to the Patrol’s (expensive) adventure with high-tech, something that continues to this day. While I believe in walls, sensors, helicopters, and night scopes, sometimes I think the Patrol goes too far. It tries to develop techie solutions to enforcement challenges that are not sensible uses of scarce federal dollars and sometimes seems to be playing over its head.

Drones. The Border Patrol has not done well with drones, for instance. A few years ago, my colleague Dan Cadman wrote, quoting from a long DHS Office of the Inspector General report that:

"The CBP has already spent $360 million on the drone program over the past eight years, breaking down to about $12,000 per [operating] hour — and yet has proven ineffectual for the 150 miles of border being surveilled with [that equipment]."

That may have changed since. In this situation, we have a relatively easy challenge in that everyone in the world is using drones for things as mild as package delivery to things as lethal as government-directed assassinations; the Border Patrol needs to figure out a way to use the drones’ capabilities to meet its own needs.

Tunnels. In contrast, there is a need for up-to-date ways to detect the digging of tunnels under some parts of the southern border. Such tunnels are designed more to move drugs than illegals, but both are Border Patrol challenges. Tunnel detecting technology is — as opposed to drones — a subject of interest to exactly two customers: the Border Patrol and the government of Israel. There is not as large a body of DoD research and development on tunnels as there is in the case of drones.

Fortunately, this is not a problem east and south of El Paso, as the Rio Grande prevents (or has so far) the digging of tunnels under it. But west of that city, there are some built-up urban areas that have seen the cartels build numerous tunnels — some with interior lighting systems, forced air, and rudimentary railroad tracks. To my knowledge, no generally useful tunnel-detecting technology has been developed.

One low-tech way of detecting tunnels, sometimes used by Mexico’s police forces, is to notice what happens to the large amounts of spoil, all the dirt and rocks that must be disposed of to keep the tunnel moving forward. This is usually done at night. A low-tech reward system for information on dump trucks moving away from the border at night might do the trick.

Mechanical Dogs? But sometimes the Border Patrol seems to be looking for solutions to rare or nonexistent problems, a thought I had when reading a very enthusiastic DHS press release about its work creating robot dogs.

DHS calls them automated ground surveillance vehicles (AGSVs). Why not DOGS, for designated operational ground surveillors?

They are supposed to check out situations that real dogs or human beings cannot handle because of danger. DHS is vague about just what the robots would or could do. The press release is full of copy like this:

Later, in a desert area, the [robotic] dogs were programmed to go on simulated sentry duty. Under this autonomous mode setting, the AGSVs headed out and made turns when they reached pre-determined GPS waypoints. After completing their circuit, they returned to base. This was done in the daylight, as well as at night. Additional testing included putting the dogs through the paces of simulated inspections outside, inside, and under train cars at railyards.

There is no information on the cost of these dogs.

As I said, a sexy solution in search of a problem.

Why not spend the money on ferrying illegal aliens back to their homelands?