Laura Bliss is a staff writer at CityLab, covering transportation and technology. She also authors MapLab, a biweekly newsletter about maps (subscribe here). Her work has appeared in the New York Times, The Atlantic, Los Angeles magazine, and beyond.
Santander, Spain, tested a network of acoustic sensors capable of managing traffic congestion. But will it stand up to the future of cars?
Los Angeles traffic and law enforcement tried, and failed, to stop the Olympic cyclists from racing on the freeway. And so on Sunday, August 5, 1984, a 17-mile stretch of southern California's Route 91 was closed to traffic for 16 hours as long-distance bikers pedaled for the gold.
It was just one of many traffic-related headaches the '84 games caused L.A. Yet the anticipation of Olympic congestion also inspired the city's famed Automated Traffic Surveillance and Control (ATSAC), a system of just-below-pavement electromagnetic coils that pick up on passing cars and transmit traffic data to a central computer, which enhances "green-time" to ease backups. The system proved so effective in stadium-adjacent tests that it's since expanded to some 4,400 intersections around the city. It also inspired many U.S. metros (Buffalo, Pittsburgh, Portland, and Atlanta, to name a few) to develop similar congestion-control systems based on magnetic induction, video data, radar, or even car Bluetooth signals.
But a new traffic-easing technology relies on a different form of stimulus: Sound. Last month, researchers Pedro Malo of Portugal's UNINOVA institute and Philippe Cousin of IT consulting firm EGlobalMark completed a two-year trial of Ear-It, an experimental network of outdoor acoustics sensors, as part of the Spanish city of Santander's recent "smart city" initiative.
Several hundred acoustic processing units (APUs) capable of measuring traffic volumes, speeds, vehicle-lane occupancy, parking availability, and "sonic events" like sirens and gunshots were installed around the city and connected to a central processor. Embedded with microphones and tiny computers equipped to run sound-interpreting algorithms, the APUs worked in tandem with a number of cheaper "motes," (referred to as "IOTs" in the header image), whose wider distribution helped pinpoint exact locations of sounds—useful if a shot is fired, or if a loud event is taking place.
The two-year experiment in Santander helped evolve the lab-developed technology, says Malo, testing it against weather and a range of interfering noises. "We now have a matured technology that stands up to demanding requirements," he says. "We have a lot of cities interested in using it." He points to the system's clear advantages over the old-school, electromagnetic-coil approach: It's much easier to install, with no tearing up pavement or shutting down streets. "You put it on a lamppost, and that's it," he says. And unlike other video or Bluetooth-based systems, the APUs are capable of gathering data far beyond cars counts. "The key word is really multi-purpose," says Malo.
But do acoustic sensors measure up to what many in the U.S. view as the future of traffic lights—which is no traffic lights at all? Once vehicles are uniformly equipped to communicate with one another (the federal government is taking steps to require V2V technology in new models), traffic control may shift from external infrastructure like lights and sensors to technology within cars themselves.
"My guess is that in future we're going to have all vehicles be connected, vehicle to infrastructure capabilities," says Chris Hendrickson, Professor of Civil and Environmental Engineering and Co-Director of the Green Design Institute at Carnegie Mellon University. "So in the future you could also use those signals for counting vehicles and managing congestion that way." He also points out that many ambulances are already capable of overriding traffic signals. There's been experimentation with putting connectivity tools on buses, too.
So Ear-It may face an uphill battle for traffic applications, at least in the U.S., where most metros have management centers with varying levels of capabilities. But it could find its niche in cities without an existing traffic control system—or outside of the traffic world entirely. Malo and Cousin also tested the APUs indoors, in order to detect emergencies (like an elderly person falling) and to track an office building's energy use.
Tom Neeld, a researcher at University College of London's Energy Institute, is also developing ‘intelligent acoustic sensors’ that track energy consumption of gas boilers. "For all energy-consuming mechanical devices, acoustic sensing has a major advantage in that nothing is hidden," he says by email. "Other sensors, such as light sensors, are very directional and will only tell me information about one part of the system to which it is pointing... In my work, the microphone picks up every process that is occurring within a complex system."
Malo says the same is true of Ear-It. The full results of the experiments, both indoor and out, will be posted online following a final review in late November. And for those already interested, a useful "benchmark procedure" is already available to determine whether your town is prepared to put ears to the ground.