Image: a TomTom traffic map of London.
Too much traffic
So many commercial delivery cars have hit the road in recent years that they’ve become a big cause of gridlock, reduced safety, and increased pollution. In a contributed (and most likely sponsored) piece at TechCrunch, Automotus CTO Harris Lummis makes a case that computer vision tech can act as a meaningful solution to these problems.
And he’s not the only one who believes this.
Data for better parking policies
Here’s Lummis’ argument: By placing cameras in public places to monitor the movement of vehicles, cities can feed real-world data to a computer vision solution and extract the information they need to better manage traffic.
One potential use case is to extract real-world data on the behavior of commercial delivery vehicles and private cars looking for parking spots, a huge cause of unnecessary driving and emissions. “By collecting comprehensive data around the demand for curbside space,” he writes, “cities can design parking policies that ensure proper alignment between the supply of curb space and the way vehicles are actually using it.”
During a pilot for this method, Lummis says that Automotus was able to drop traffic by more than 20%. Cars turned over faster, spent less time looking for spots, and ultimately reduced traffic and emissions.
An emerging market?
Automotus is not the only company looking to apply computer vision in cities to solve big transportation problems. In fact, this looks to become a bigger market for computer vision technology.
Allvision, for instance, combines data from GIS mapping platforms, satellite imagery, and autonomous vehicle sensors to generate “rich geospatial models.” Next, the solution applies machine learning and computer vision to this data for applications like parking management, vegetation encroachment monitoring, and time-based asset inventories.
Another company, Carmera, gathers real-time road data using consumer-grade cameras mounted on commercial fleet vehicles (the very delivery vehicles that Automotus identified as an issue). Using machine learning and computer vision, the company offers “change-as-a-service,” which automatically recognizes changes to the road, from new signals to temporary events like construction.
To learn more about Lummis’ big plans for computer visions in cities, read a few more use cases, or check out some more statistics, click through to the article here.