Cities leverage new technology to find & fill potholes
Local governments are turning to tools like AI and machine learning to make pothole repair more efficient, more environmentally friendly, and most cost-effective.
A number of local governments, such as Richmond Hill, ON and the London Borough of Brent, GB, are using artificial intelligence (AI) to review imagery of roads and identify potholes without a physical
In Memphis, TN the city outfitted city busses with cameras to capture road footage that was then put through an AI application.
Tarrant County, TX is leveraging reports from the mapping and traffic app Waze to then auto-generate a work order and dispatch a crew to fix the damaged roads.
Kansas City, MO introduced a slew of pothole-related initiatives, including a rapid-patch that uses gel-based bags to fill holes temporarily until the city is able to permanently fix it.
Anyone with a car has likely experienced that hold-your-breath-and-brace-for-impact feeling when your wheel hits a pothole. Even if your car passes over the bump unscathed, it’s a frustrating and potentially dangerous experience.
For cities, these divots in the road are more than an annoyance. They’re a multi-million-dollar problem.
Potholes are part of the lifecycle of a road, and finding and fixing potholes is a regular part of a local government’s road maintenance system. But while cities may not be able to fully prevent potholes from happening in the first place (yet), they can enact smarter ways to manage them. A number of municipalities around the world are leading this charge, $ employing new technology$ such as AI and machine learning to identify and prioritize potholes that need to be fixed, and leveraging new materials and techniques to get them filled.
The Problem with Potholes
Potholes form when water seeps into the soil below a road. Over time, the road surface breaks down and a hole forms in the pavement. These cavities can form at any time — the simple act of cars regularly driving on a road can do the trick — but winter tends to be a particularly strong culprit, as water in the ground freezes and thaws and weakens the asphalt.
Once there, potholes can cause a lot of problems. There’s the potential for car damage — likely the most familiar risk to anyone who’s owned a car. $ The American Automobile Association estimates that potholes cost American drivers about $3 billion a year in car repairs$ .
Beyond damage to vehicles, they can also cause accidents, resulting in injury or death. Then there’s the degradation of road quality. Cities often set aside a $ dedicated budget for potholes$ (though regional and national governments also cover some of the repair cost). For example, Edmonton, AB, whose snowy winters unfailingly stress the roads, manages nearly 500,000 potholes a year. The city $ recently added 5 new pothole-filling trucks to its fleet at a cost of $400,000 a piece$ .
Many communities allow residents to report potholes via 311 and other citizen engagement apps, but these still need to be assessed and patched by city workers. This uses up local government staff resources, and because it requires vehicles to drive and repair the roads, the process also can add to local emissions.
So how can local governments streamline the process, when potholes themselves don’t appear on any schedule?
Technology can help. A number of cities have been experimenting with innovative ways to make it easier and more efficient to repair potholes. Let’s take a look at what they’re doing.
Innovative ways to identify potholes
In order to fix potholes, you first have to find them. Here’s how smart cities are doing this:
Richmond Hill, ON uses AI and a dashboard smartphone
In $ Richmond Hill, ON$ , the $ city piloted ROVER, a system that uses AI$ to identify, measure, and catalogue potholes from images captured on a dashboard-mounted smartphone. Traditionally, Richmond Hill potholes are reported by residents, and then the city conducts manual inspections to assess the damage. In 2019, 1,078 potholes were identified and fixed this way.
$ ROVER$ , which is developed by Richmond Hill-based company Visual Defence, takes the work out of the pothole analysis. A smartphone films the road as a vehicle drives, and the app uses AI to mark, measure, and log the address and GPS coordinates of potholes along the way, with a 95-percent accuracy rate. $ Over the course of a 135-day pilot, ROVER helped the city identify and repair 542 potholes$ . Richmond Hill is now rolling out ROVER to $ assess other road damage problems$ .
Memphis, TN turns city busses into pothole scouts
Like Richmond Hill, $ Memphis, TN$ , is also using AI to take the manual work out of identifying potholes.
The city was spending a lot of time and money to fix potholes every year. In one recent year, Memphis crews repaired 63,000 potholes, which took about 32,000 worker hours. Only 20 percent of these potholes were reported by residents — and those were usually the ones that had become bad enough to make people unhappy. The rest of the repairs had to be identified by street maintenance crews, which, given Memphis’s large size, was no small feat.
$ Robert Knecht, Public Works Director for the City of Memphis, summed up the problem$ :
"Historically, Public Works has relied primarily upon Street Maintenance crews to proactively locate and fill potholes. As Memphis has over 6,800 lane-miles of public streets, it is a daunting task to reliably survey the entire system in an efficient and systematic way.”
So Memphis $ partnered with Google and SpringML to use AI to analyze videos of road conditions and identify potholes$ . To shoot the video, Memphis equipped city busses with cameras — as the busses criss-crossed the city, they captured miles of road footage. The city then layered in data from 311 reports to help refine the model, until the AI application was able to tell the difference between a pothole and other objects, like manhole covers.
The project was a big success, with an estimated 75-percent increase in potholes detected. This is expected to $ save the city around $20,000 a year$ . On top of this, by using busses, the city is able to record and review road quality using existing vehicles and routes, without placing a lot more vehicles on the road and driving up emissions.
Tarrant County, TX repurposes a popular traffic app to find potholes
Memphis made pothole identification easier by using city busses. In $ Tarrant County, TX$ , it’s $ Waze that’s working to find the damaged streets$ .
The county, which owns 415.7 miles of roadway across 4 precincts, has $ a Waze dashboard$ that aggregates reports from users of the mapping app, who report road hazards such as potholes while on the road. Using APIs and two-way data sharing, the county has connected the data from $ Waze$ with $ Cartegraph$ , an operations management system. When a pothole is reported, Cartegraph automatically sends a work order to the appropriate field maintenance team.
$ There are multiple benefits, as documented by Tarrant County$ :
“Automating crowd-sourcing information reduces the need for a general crew or inspector to navigate the roads searching for situations needing repair, documenting, and re-dispatching a specific crew to address the need. Additionally, Tarrant County reduces precinct fuel costs, saves staff hours, and reduces response times.”
Repairing potholes more efficiently
Of course, once a city identifies a pothole, the next step is to fix it. While thi still usually includes some degree of manual effort, there are a lot of ways that cities can streamline their workflows and use technology to make the process run more smoothly.
London Borough of Brent, GB partners with AI to find and fix potholes
As the pandemic kept more people at home, the London Borough of Brent, GB, decided to take advantage of the less-congested roads. The $ borough partnered with Vaisala, an environmental measurement firm, and Yotta, an asset management platform$ , to undertake an innovative new way to identify and repair potholes around the city.
Over the course of just six days, a dash cam-equipped vehicle drove all 340 miles of the borough’s streets. The video was then analyzed using AI and identified 4,900 potholes. The Velocity technology not only logged the holes, it prioritized them according to what needed to be fixed most urgently.
From there, the Brent Council began patching the holes based on the prioritization. The borough used the $ Velocity method$ , which reduces carbon emissions by 90 percent and takes a fraction of the time to complete. The entire process cost half of what the borough’s traditional pothole-filling process totaled.
According to $ Cllr Krupa Sheth, Lead Member for Environment$ :
“By rethinking old approaches and exploring new technologies, we’ve been able to repair more roads at the same time as saving money and cutting our carbon footprint. I’m thrilled with what we’ve been able to deliver for residents in such a short amount of time.”
Kansas City, MO uses new gel technology to rapidly fill holes
The easier it becomes to zero in on potholes, the more quickly cities will have to address them. In $ Kansas City, MO$ , a new tactic is helping the city patch holes as it finds them — temporarily at least.
Potholes complaints have been on the rise in Kansas City; $ by February 2020 the city had already received over 1,500 complaints$ for the year. What’s more, the city has $ reimbursed around $125,000 in claims$ filed by residents whose cars have been damaged by the holes over the course of a year.
To try to keep up, the $ city is using “thump pads,” which are reusable bags filled with gel $ that can be placed in a hole to minimize the impact on cars until a more permanent filling can be added. Weather not ideal for laying asphalt? Toss in a thump pad. Truck and crew not available? The pads can be put down without heavy machinery.
$ Maggie Green, a Kansas City spokesperson, explained the new technique$ :
"It's a super fast thing that you just throw out. You find the pothole, you put the thump pads in and you can drive off without having to mess with a big asphalt truck."
Thump pads aren’t the only innovation Kansas City is rolling out to fix its roads. The city has experimented with $ using technology to predict potholes$ . They publish interactive maps. The $ mayor even appointed a “pothole czar”$ to focus on street repairs (though the concept has met with resistance within the city council).
Potholes themselves aren’t going away any time soon. As long as roads are made of asphalt and subjected to elements like rain and snow, they’re going to erode over time.
But this doesn’t have to be a losing battle for local governments trying to maintain their streets. By adopting new technologies that make it easier to stay on top of the issue, cities and counties can identify pothole problem areas sooner. They can work more efficiently and prioritize repairs. And they can do all this while saving money and in some cases, reducing emissions.