America’s road infrastructure carries an alarming grade of D+ from the American Society of Civil Engineers, a rating that reflects decades of underinvestment and deferred maintenance. Identifying which roads need repair has traditionally relied on subjective visual inspections and costly specialized surveys. The RoadBotics 2021 U.S. Cities Road Report introduced a paradigm shift by using windshield-mounted smartphone cameras and machine learning to assess 75 miles of roadway across 20 cities, producing objective, comparable road condition data. For contractors and public works professionals, understanding how this technology works and what the findings reveal is essential for making informed infrastructure investment decisions. For those managing quality documentation, a Non Conformance Report Ncr How to Report Construction provides a structured format for documenting pavement condition deviations.
The Road Condition Assessment Gap: Why Traditional Methods Fall Short
The ASCE Infrastructure Report Card gave the United States an overall C- across roads, bridges, ports, drinking water systems, and electricity grids, identifying a $2.59 trillion spending gap. The roads category earned a D+, reflecting widespread deterioration affecting safety, travel times, and vehicle operating costs. At the time of the report’s release in early 2021, U.S. Transportation Secretary Pete Buttigieg faced the task of overseeing over 4 million miles of American roads, demanding reliable data on actual pavement conditions.
The Limitations of Subjective Pavement Inspections
Traditional road condition assessments have significant drawbacks:
- Manual visual surveys require trained inspectors to drive road segments and assign ratings. Results vary between inspectors and cannot be reproduced at scale.
- Professional pavement management systems use specialized vehicles with laser profilers and ground-penetrating radar. While accurate, these cost tens of thousands per mile and are impractical for smaller municipalities.
- Public complaint-based maintenance relies on citizen reports to find problem areas. This reactive approach catches only the worst roads and misses deteriorating segments before they become hazards.
- Aggregated state data compresses local variability into regional averages, hiding the fact that one city may have excellent roads while a neighboring jurisdiction struggles with widespread failure.
RoadBotics CEO Ben Schmidt noted that when Buttigieg was mayor of South Bend, Indiana, he brought RoadBotics to the city to obtain an objective assessment of the entire 550-mile road network before spending a dollar on repairs. This approach should serve as a model for national infrastructure decision-making.
The Cost of Not Knowing
Without objective condition data, infrastructure funding becomes a political negotiation rather than an engineering decision. Schmidt stated that a new administration presented an opportunity to close a projected funding gap of more than $2 trillion needed by 2025, but meaningful allocation requires knowing the exact state of the nation’s infrastructure. He emphasized that objective condition data is an essential starting point for managing infrastructure, and that data-driven decision-making is the only way to improve America’s infrastructure grade.
For construction professionals assessing site conditions, the Understanding a Dilapidation Report in Construction provides a framework for documenting pre-existing conditions that helps contractors avoid disputes over responsibility for pre-existing damage.
Inside the RoadBotics Assessment Methodology
RoadBotics conducted its 2021 U.S. Cities Road Report by assessing 75 miles of roadway across 20 US cities, covering a representative sample of urban road networks in different regions. The methodology combines smartphone data collection with machine learning to create a repeatable, objective assessment system.
How the Technology Works
The assessment process follows a technically rigorous workflow:
- Data collection via windshield-mounted smartphone. A standard smartphone is mounted on a vehicle’s windshield and captures continuous images of the road surface at regular intervals. No specialized vehicles or laser scanners are required.
- Image processing and feature extraction. Computer vision algorithms identify key pavement distress indicators: cracking patterns, surface deterioration, rutting, and potholes. The system distinguishes between different crack types, severity levels, and surface defects.
- Machine learning classification. A trained model analyzes the extracted features and assigns each road segment a condition rating on an objective scale, trained on thousands of labeled road images.
- Comparative scoring across cities. Because all cities are assessed using the same technology and scoring criteria, the results are directly comparable, eliminating the subjectivity of traditional inspection methods.
Key Advantages Over Traditional Methods
| Assessment Method | Cost per Mile | Objectivity | Scalability | Typical Users |
|---|---|---|---|---|
| Manual visual inspection | Low ($200-500) | Low (subjective) | Moderate | Small municipalities |
| RoadBotics smartphone method | Very low ($50-150) | High (ML-driven) | High | 300+ governments |
| Specialized laser profiling van | High ($2,000-$5,000) | High | Low | State DOTs, large counties |
| Ground-penetrating radar | Very high ($5,000-$15,000) | High | Very low | Research, major highways |
Key Findings from the 2021 U.S. Cities Road Report
The report marked the first time RoadBotics technology was used to create a comparative analysis of road condition ratings across select US cities. The objective nature of the data allows for easy comparison to identify where road networks need the most help.
Pervasive Deterioration Across All Regions
No city in the survey received a uniformly excellent rating. Even cities with relatively strong road networks had significant pockets of deterioration, highlighting that America’s D+ ASCE grade reflects a nationwide problem, not one confined to particular regions.
Variability Between Neighboring Jurisdictions
One of the report’s most striking findings is the significant variability between cities that share similar climates, economies, and demographics. This suggests that local maintenance policies and investment decisions play a larger role in road quality than previously understood. Factors that appeared to influence condition ratings included:
- Pavement age and original construction quality. Roads built to higher specifications showed less deterioration even under similar traffic loads.
- Maintenance history and timing. Cities performing routine crack sealing at the right intervals had significantly better ratings than those letting small defects propagate.
- Freeze-thaw cycle exposure. Northern cities showed more rapid deterioration, moderated by drainage quality and appropriate pavement mixes.
- Traffic loading patterns. Roads on truck routes deteriorated faster, though the gap was smaller in cities with effective load management.
The Data Gap Hinders Funding Allocation
A critical finding of the report is that the lack of objective, comparable data across jurisdictions prevents efficient allocation of infrastructure funding. Schmidt argued that without knowing the exact state of the nation’s infrastructure, policymakers cannot have a robust debate on the most appropriate way to allocate infrastructure stimulus packages. This data gap was particularly relevant as Congress debated major infrastructure investment legislation. Buttigieg, during a presentation of the ASCE infrastructure report card, stated that the country had a long way to go and that infrastructure was in tough shape. He characterized the moment as a once-in-a-lifetime opportunity to invest, adding that the country was past the point of allowing Infrastructure Week to remain a Washington punchline.
Applying Data-Driven Assessment to Infrastructure Projects
The principles behind the RoadBotics report have direct applications for construction professionals and public works departments. Moving from reactive, subjective assessments to proactive, data-driven approaches improves outcomes, reduces lifecycle costs, and supports defensible budget requests.
Practical Steps for Objective Pavement Assessment
- Baseline your existing network. Conduct an objective condition assessment before committing to repair strategies. Smartphone-based methods make this feasible even for smaller jurisdictions.
- Segment roads by condition severity. Divide the network into categories from preservation-only to full reconstruction, allowing targeted budget allocation.
- Match treatment to condition. Early-stage cracking responds to sealing; advanced deterioration may require milling and overlay; failed roads need full reconstruction.
- Track changes over time. Repeat assessments annually to measure maintenance impact and detect accelerating deterioration before it becomes expensive.
- Use data for stakeholder communication. Objective data supports more convincing budget presentations to elected officials and the public.
Why Data Matters for Contractors
For contractors bidding on road repair and reconstruction projects, access to objective condition data has several practical benefits. It enables more accurate bidding by allowing contractors to estimate work required more precisely, reducing the risk of cost overruns. Data showing the deterioration trajectory of different road segments helps contractors advise clients on which repairs to prioritize. Post-construction condition assessments using the same technology provide an objective baseline for warranty periods and long-term performance monitoring, reducing disputes between contractors and owners. Contractors who use data-driven approaches can demonstrate to municipalities that they understand the full scope of infrastructure needs, positioning themselves as strategic partners rather than transactional service providers.
The broader context of walkable urban infrastructure is increasingly relevant to roadway planning. The Walkable Cities Report 2025 Key Findings Every Home explores how road condition and design affect pedestrian accessibility and community livability.
Connecting Road Assessment to Site Preparation
Data-driven road condition assessment connects to a broader ecosystem of site investigation activities. Understanding existing road conditions informs decisions about material haul routes, traffic control, and the scope of pavement restoration after utility work. The Comprehensive Guide to Steps in Preparing Site for outlines how soil reports and excavation planning integrate with pavement condition data.
The Path Forward for Data-Driven Infrastructure
The RoadBotics 2021 U.S. Cities Road Report demonstrated that affordable, objective road condition assessment is a current capability, not a future aspiration. With over 300 governments already using smartphone-based pavement assessment, the technology has moved from experimental to operational. The challenge now is scaling adoption to close the data gap that prevents efficient allocation of the trillions of dollars needed to improve the nation’s infrastructure grade from D+ to something approaching adequate. For construction professionals at every level, from municipal public works directors to private contractors, the lesson is clear: objective data enables better decisions. Whether assessing a single project site or an entire city road network, investing in data collection before committing to repair strategies produces better outcomes, lower lifecycle costs, and more defensible budget justifications. The smartphone-based approach pioneered by RoadBotics has made this level of analysis accessible to organizations of any size, democratizing a capability that was once reserved for well-funded state transportation departments.