What does it take to make a city smart? According to conventional thinking, it means installing thousands of sensors and cameras everywhere to measure traffic and pollution and movement. But it turns out that some existing infrastructure could act as a sensor already–no new tech needed.
A new data visualization shows how about three miles of fiber optic cables, buried below Stanford’s campus, can detect the unique vibrations created by people driving, biking, and walking. Normally these readings are filtered out as noise since the cables usually used for sending messages from one place to another. But Eric Rodenbeck of San Francisco-based Stamen Design realized that this “noise” doubled as a way of tracking the movement of vehicles and people.
“If you treated the noise as signal, you could start to measure the difference between a Prius and a Subaru,” Rodenbeck says since the cable can measure to such accuracy that it can differentiate between the two types of cars. “I got captivated by that idea that almost accidentally we have this gigantic measuring network.”
The visualization, called “Big Glass Microphone,” takes 10 minutes of data from the cable, which is buried six inches below the ground in the shape of a lopsided figure-eight. When mapped, it’s easy to see larger moving waves of vibrations (cars), static pools of vibration (a large fountain on campus), and smaller waves (people and bikes). Commissioned by the Victoria & Albert Museum in London, the visualization will be part of an exhibition at the museum in 2018.
The cable itself is part of the Stanford Exploration Project, where scientists are using the cables to detect seismic activity and create 3D and 4D images of fault lines, like the San Andreas Fault. But you’re probably more familiar with fiber optic cables being used to transport the internet–the U.S. alone has an estimated 35 million miles of fiber optic cables crossing the country. That’s one big sensor.
Some companies are already leveraging this kind of technology, using cables as sensors instead of using them to send signals. Rodenbeck says that Optasense, which provided the data for the visualization, is installing fiber optic cable around oil pipelines and perimeters of secure areas to act as a means of detection and measurement.
This concept of using already-built infrastructure as a sensor applies to more than these cables, though. For one, Rodenbeck says that cell tower infrastructure could be used to make detailed weather maps of humidity–because the time it takes for a cell phone to communicate with a nearby tower is directly related to the moisture in the air. “This incredibly elaborate sensing mechanism just sort of falls off the truck,” he says.
And it’s accurate, too. Rodenbeck says that fiber optic sensing can even detect how many people are in a car. It could be used to monitor carpool lanes on freeways as well as the flow of traffic. If different cars also cause slightly different vibrations, then theoretically you could track which type of car is driving by as well. It’s a sophisticated system that’s already in place; it just needs to be activated. “Basically anything you can clamp a computer to and send a signal to starts to be used to measure things,” Rodenbeck says. “That’s kind of wild.”
This isn’t a reality just yet, but it reveals that making cities smart doesn’t necessarily mean embedding sensors and cameras everywhere, an expensive and time-consuming endeavor.
Unsurprisingly, such technology raises plenty of privacy concerns, since it might be possible to use each person’s unique vibrations to track their location (though many carry smartphones, another infrastructure system already in place with the potential for mass tracking).
“It’s simultaneously amazing and terrifying,” Rodenbeck says. “My most common emotional response these days is vertigo. It opens up a new set of possibilities, both positive and negative.”