Pavement management is not a software purchase. It is five connected habits — know what you own, know its condition, know how it ages, decide what to fix, and show the results. Here is how a small agency builds each one.
1. Build the inventory before you buy anything
Every program starts with a segment-level list of roads: centerline length, width, functional class, surface type and ownership. Most agencies already have 80% of this in a GIS layer or a public works spreadsheet.
Do not wait for perfect data. Load what exists, then correct it as crews and inspectors touch each segment. An imperfect inventory that gets used beats a perfect one that is still being built.
- •Start from your existing GIS centerline or street list
- •Split segments at intersections or logical maintenance limits
- •Record surface type — it drives every deterioration assumption later
- •Flag roads you do not own so they stop appearing in your backlog
2. Choose one condition method and stick to it
Condition scoring only works if it is repeatable. ASTM D6433 PCI is the most widely recognized method, but a consistent 1–5 or good/fair/poor scale beats an inconsistent PCI.
Collection can be walking surveys, windshield surveys, automated vans or drone imagery. What matters is that the same distresses are judged the same way each cycle, and that every score carries a date.
- •Pick one scale for the whole network
- •Record the inspection date and inspector with every score
- •Re-inspect on a cycle — typically every 2 to 3 years
- •Photograph representative distresses so scoring can be audited
3. Model how pavement ages
Pavement does not fail linearly. A road spends years in good condition, then drops quickly once water reaches the base. That curve is why preventive treatment on good roads is cheaper than reconstruction of failed ones.
Deterioration curves per surface type let you project the network forward and see which roads are about to fall off the cliff — the ones where a cheap treatment now avoids an expensive rebuild in four years.
4. Prioritize by value, not by worst-first
Worst-first is the default in most agencies and the most expensive strategy available. It spends the whole budget reconstructing a handful of failed streets while the rest of the network quietly slides.
Optimization ranks candidate treatments by condition gain per dollar, subject to your constraints — funding caps, district splits, must-do commitments. The output is a project list you can defend line by line.
- •Match treatments to condition: crack seal, seal coat, thin overlay, mill and fill, reconstruction
- •Use local unit costs, not national averages
- •Apply hard constraints first, optimize the remainder
- •Keep the reasoning attached to each selected project
5. Report in the language of the decision-maker
Council members do not fund PCI. They fund outcomes: how many miles improve, what happens if funding stays flat, what deferring a year costs.
Run at least three scenarios before every budget cycle — do nothing, current funding, and target condition — and bring the projected network condition for each. The City of Coachella used this approach to produce a defensible multi-year plan and justify a request for a fourfold increase in its pavement maintenance budget.
Frequently asked questions
- How long does it take to build a pavement management program?
- An agency with an existing road inventory can be running scenarios in a few weeks. Field condition collection is usually the longest step and is often contracted to a data collection partner.
- Do we need PCI to start?
- No. Start with any consistent condition scale you can collect reliably, and move to PCI when you have the capacity or a collection partner.