Here’s how AI-driven RoadBotics celebrated National Infrastructure Week

It’s no secret that infrastructure in the US is crumbling. But this Pittsburgh startup is hoping to do something about it.

Carnegie Mellon University spinout RoadBotics shared a new demo and highlight of its customers last week in celebration of the 10th annual Infrastructure Week by nonpartisan coalition United For Infrastructure.

The venture-backed RoadBotics, which was founded in 2016, specializes in using artificial intelligence to map the status and condition of a region’s infrastructure through visual and other data inputs. So far, its signature AI platform RoadWay has provided over 250 governments across the world with new assessment and management capabilities for their roads, per the company.

In honor of 10 years of Infrastructure Week, which gives focus to infrastructure as a critical issue, RoadBotics highlighted 10 case studies it’s performed with 10 different customers using its AI-powered mapping technology. It also hosted a live virtual demo of its technology, the latest in a series of demos the company’s been sharing recently, with past recordings available here. The next live demo is coming up on June 8, with registration available here.

Those are worth checking out to get a sense of just how RoadBotics’ tech works for its customers, which are spread across 34 countries worldwide and 34 US states. But in the meantime, here are some local communities the company highlighted last week, and case studies on how RoadBotics has helped them improve their infrastructure:

Cranberry Township

Just north of Pittsburgh, Cranberry Township formerly used a platform called Roadsoft to evaluate the area’s thousands of road segments according to a PASER rating scale that was determined by the streets and fleet manager for the township. That process was highly subjective, a case study from RoadBotics found.

“It’s a challenge to send one person out regularly to use their best judgement and determine the rating of the road conditions,” said Jason Dailey, former public works director of Cranberry Township in the case study. “You’re always allowing human error to influence the ratings.”

By implementing RoadBotics’ products instead, the township had an objective AI-powered way of rating the roads, with visual data from the mapping process to back it up.

Bethel Park

Similar to Cranberry Township, Bethel Park (a suburb south of Pittsburgh) also had an inefficient and subjective way of evaluating their roads. For two weeks each year, engineers would go out and assess the 118-mile road network in person. Then, the engineers would enter the information into two asset management platforms, neither of which truly met their needs for analysis or user interface.

Enter RoadBotics, which provided its first assessment to Bethel Park in 2017. The case study on that assessment and those that followed shows that RoadBotics’ approach was again more efficient and objective than previous methods.

“We’re very satisfied with the assessments and would 100% recommend RoadBotics to every municipality to help them with their road maintenance,” Bethel Park Municipal Engineer Lou Lambros said in the report.

Peters Township

Peters Township, another suburb south of Pittsburgh, struggled with the same challenges that Cranberry Township and Bethel Park had — their road infrastructure evaluation methods were slow, and the upkeep and preventative measures taken on small problems was inefficient. In other words, if a road had a small problem, Peters Township’s approach would delay maintenance on it in order to address more serious problems, and typically wait until the small problem of the initial road became big enough to demand urgent repaving.

RoadBotics launched its first assessment of Peters Township in 2020, enabling the government to use smart phones on the dashboards of cars that would drive around and collect visual data on the township’s roads.

“Peters Township was able to objectively determine and confirm the worst 25 miles of roads using RoadBotics 1 – 5 rating system, where 1 is dark green, representing the best roads, and 5 is red, noting the worst,” the case study reported. “After reviewing and sub-ranking those initial 25 miles, they used the assessment to create a new three-year plan. The assessment provided objective feedback on their decisions, and the online platform allowed them to perform some aspects of their planning remotely.”

North Huntingdon

Finally, RoadBotics shared a case study on North Huntingdon, a township southeast of Pittsburgh, in its roundup of 10 use cases for National Infrastructure Week. Before teaming up with RoadBotics in 2017, North Huntingdon’s township carried out a process that took nearly six months to fully evaluate the road conditions on its 160-mile network, often disrupted by weather and limited resources.

RoadBotics’ helped complete that same process in the total span of one month, including the time needed to triage road repairs and make plans to start them.

“Our engineer was spending less time out in the field and only looking at the refined list of roads in the worst condition that we generated from RoadBotics data,” North Huntingdon Associate Planning Director Ryan Fonzi said in the case study. “So, it was huge for time-saving and cost-saving as far as labor costs.”

Sophie Burkholder is a 2021-2022 corps member for Report for America, an initiative of The Groundtruth Project that pairs young journalists with local newsrooms. This position is supported by the Heinz Endowments.