Good data means money well spent for Morgan Musick, an engineer at the Alabama Department of Transportation. High on her wish list is reliable, up-to-date information about the condition of every guardrail, road sign, and median strip marking Alabama’s 11,000 miles (about 17,703 km) of state roads and highways. The data needs to be so good that she can stake her department’s budget on it.
Alabama is among the first U.S. states to adopt performance-based budgeting for road maintenance. The approach allocates money based on what the roads actually need, rather than on what each district requested the previous year. But the execution has been far from straightforward. The approach depends on good data, and for most of the past 15 years, that data came from crews driving the network of roadways with clipboards, cameras, and practiced eyes.
The results were valuable but uneven. Different inspectors saw different things. Coverage varied as some stretches of road got surveyed more often than others. “To strengthen our performance-based budgeting, we need consistent, quantified data to produce condition assessments across all districts,” said Musick, the department’s assistant maintenance management engineer.
The solution Alabama chose is Blyncsy, a platform powered by artificial intelligence (AI) from Bentley Systems, the infrastructure engineering software company. Blyncsy turns crowdsourced dash-camera footage into a machine-readable rendition of a road network. Cameras mounted in fleet vehicles and commercial trucks capture high-resolution imagery as they drive. The footage uploads to the cloud, where it’s anonymized by the camera provider. Blyncsy then feeds the anonymized imagery to its computer vision algorithms.
The AI algorithms can identify and assess more than 50 types of roadway assets and conditions — from cracked guardrails and obscured signs to fading pavement markings and vegetation encroaching on shoulders. A pilot project demonstrated 97% accuracy, the kind of number that earns the trust of engineers who build careers on precision.
Alabama’s adoption of the Bentley technology, announced in early February, is not happening in isolation. It is the latest in a series of state transportation departments turning to Blyncsy to solve a version of the same problem: how to clearly and consistently see an entire road network, and use the information to make sound decisions that benefit residents and create safer roads.
From Aloha to the Gulf coast
The Hawaii Department of Transportation also recently announced an inventive deployment of the same technology. The state’s “Eyes on the Road” program, developed in partnership with the University of Hawaii, is distributing 1,000 free high-resolution dash cameras to residents across the lush, volcanic islands. Traditional inspection fleets can’t cover every mile often enough. But volunteer drivers become rolling sensors. As they commute, the cameras capture road conditions — guardrail damage, debris on shoulders, vegetation blocking sightlines — and upload the imagery to the cloud for Blyncsy’s AI to analyze.
“Our crews have been working with Blyncsy to refine the machine learning algorithms to amplify our efforts to efficiently maintain our transportation infrastructure,” said Ed Sniffen, director of the Hawaii DOT.
Alabama’s landscape is different, but the math is similar: an efficient budgeting system is only as good as the data underneath it.
The Pattern behind the projects
Blyncsy’s trajectory reads like a case study in how infrastructure technology scales. The company was founded in 2014 in Salt Lake City by Mark Pittman, who has described his motivation as frustratingly personal: a broken traffic light near his home that took months to fix because no one in government knew it was broken. Bentley acquired Blyncsy in 2023, folding it into an asset analytics portfolio that already included OpenTower iQ for telecom towers.
It was a timely deal. A federal deadline is now accelerating the turn to technology. By September 2026, the Federal Highway Administration requires all state and local transportation agencies to adopt a method for maintaining minimum pavement marking retroreflectivity, a term that describes how well lane markings reflect headlights back to drivers. Faded markings contribute to roadway departure crashes, one of the leading causes of traffic fatalities in the United States. Meeting that standard means knowing the condition of markings across an entire roadway network.
Manual measurement is possible, but it’s slow and expensive. Blyncsy released a free public map scoring retroreflectivity for roads across all 50 state capitals, giving every local department of transportation a starting point. For agencies that need network-wide coverage and ongoing monitoring, the platform offers a path to compliance that doesn’t require doubling the agency’s inspection staff. “The future of infrastructure asset management depends on making financial decisions based on empirical evidence rather than historical precedent,” said Pittman, now senior director of transportation AI at Bentley.
What makes the Alabama and Hawaii stories more than technology announcements is that they show how governments are starting to rethink their approach to infrastructure. Both states already had maintenance programs. Both had data. What they lacked was consistency — the ability to compare the condition of two guardrails on opposite sides of the state, using the same standard, measured the same way, at the same frequency.
That consistency turns data into insights and better decisions. Alabama can now tie condition scores directly to budget allocations, making the case for every dollar empirical rather than anecdotal. Hawaii can detect a hazard on a remote Maui highway hours after it appears, not months.
States adopting AI for road maintenance are not the ones with the biggest budgets or the newest highways. They are the ones that decided the gap between what they know about their roads and what they need to know had become too expensive. With a federal deadline approaching and the technology proving itself across geographies and government departments, the question for adoption is becoming less “whether” and more “when.”
