Pavement Condition Viewer

Pavement Condition Viewer

Inspiration

Cities collect detailed infrastructure data, but it is rarely presented in a way that supports clear, transparent decision-making. When we saw that Cyvl had given us access to detailed pavement, asset, and imagery datasets, we saw an opportunity to turn that raw data into something practical and understandable. Our goal was to build a tool that helps people quickly understand infrastructure conditions and identify where repairs should be prioritized.

What it does

Our Pavement and Sidewalk Viewer transforms Cyvl’s infrastructure data into an interactive decision-support tool. Users can explore pavement condition scores (PCI), view sidewalk condition ratings, and instantly see which segments are in the worst shape.

Road and sidewalk segments are color-coded by severity, making condition patterns visible at a glance. A dynamic leaderboard ranks the lowest-performing segments to highlight high-priority repair areas. When a segment is selected, users can view a 360° street panorama to visually confirm on-the-ground conditions. The platform also provides live distribution summaries so users can understand how condition levels are spread across the visible area.

For each selected segment, the platform also generates a planning-level repair cost estimate based on segment geometry, condition severity, material type, and typical Massachusetts municipal unit costs. This provides a rough order-of-magnitude (ROM) estimate to help users understand not just which segments are worst, but the potential scale of investment required.

Together, these features help translate complex infrastructure datasets into clear repair priorities.

How we built it

We built the frontend using React for state management and interface structure, and Leaflet.js for high-performance interactive mapping. Cyvl’s GeoJSON datasets were processed and filtered to support real-time rendering, category toggling, and condition-based ranking.

We implemented custom logic to classify segments by PCI severity, compute geometric properties like approximate segment length, and match map clicks to the nearest 360° panorama using distance calculations. Careful use of memoization and selective re-rendering ensured smooth performance even when displaying hundreds of segments simultaneously.

We also implemented a lightweight cost estimation model that infers likely repair treatments from PCI and condition categories, calculates material quantities from geometry, and applies planning-level municipal unit cost ranges to produce rough investment estimates.

Challenges we ran into

Rendering large GeoJSON layers while maintaining smooth zooming and panning required performance optimization. We had to carefully manage feature filtering, state updates, and map layer resets to prevent lag.

Another challenge was transforming raw infrastructure data into meaningful insight. The datasets contain many fields and formats, and we had to normalize condition values, handle missing data, and design logic that consistently categorizes severity. Ensuring that panoramas matched accurately to selected segments also required spatial reasoning and distance calculations.

Accomplishments that we're proud of

We’re proud of turning raw infrastructure data into a tool that feels intuitive and grounded in reality. The integration of 360° panoramas makes the data tangible, since beyond just seeing scores, users get to see the actual street conditions behind those scores.

We also built a system that moves beyond simple visualization by ranking and highlighting the most critical segments. The result is a tool that feels practical, transparent, and usable for real-world infrastructure discussions. By incorporating estimated repair costs alongside condition severity, the tool moves beyond visualization and begins to frame infrastructure decisions in financial terms.

What we learned

Before this project, none of us had worked directly with GeoJSON infrastructure datasets. We learned how to parse, filter, and interpret spatial data, and how to translate it into interactive map layers that update in real time.

We also gained a deeper understanding of how pavement and sidewalk condition scoring works, and how spatial data can support infrastructure prioritization. Working with real municipal-style data highlighted both the power and complexity of public infrastructure systems.

What’s next

Next, we could expand the viewer to additional cities and integrate more Cyvl datasets, including deeper asset layers and potentially LiDAR-derived insights. We could refine the cost model with more detailed municipal bid data and asset-level inspection inputs to improve estimate accuracy and prioritization confidence.

Long term, this has the potential to evolve into a broader infrastructure intelligence platform that combines condition data, visual validation, and public transparency to support smarter repair planning.