Field notes·

Field notes: the week budget season almost funded the wrong ten miles

Close-up of a rain-slicked, cracked asphalt road surface.

It was the Tuesday before the capital improvement plan went to the board, and Dale from the rating crew walked into my office holding a printout like it had bitten him.

"County Road 14," he said. "We've got it ranked third for next year's overlay. I don't think it should be in the top twenty."

Three weeks earlier, 14 had looked rough from the truck. Transverse cracking, some edge raveling near the shoulder, enough alligatoring past mile marker 6 that the rater bumped it up two condition classes on gut feel. That's not a knock on the rater. He'd been out since 6 a.m., driving a route that puts 14 right after a stretch of chip seal that makes everything after it look worse by comparison. Windshield surveys are honest work done by tired people on a schedule, and the schedule doesn't care about contrast effects.

What actually happened: a drainage issue upstream had saturated the base under about 1.2 miles of 14, and that section genuinely needed attention. The other 8.8 miles were fine, maybe five years from needing anything. But the rating sheet doesn't average down. One bad mile can pull the whole segment's score up with it if the segment boundaries were drawn by township line instead of by condition change, which is how a lot of legacy PMS data still gets cut.

Meanwhile, County Road 31, out past the quarry, hadn't been driven in fourteen months because it's low-ADT and always gets pushed to the back of the rating route. Nobody had eyes on it. It turned out to have block cracking across nearly its full length, the kind that's two seasons from needing full-depth patching instead of a mill and overlay. If we'd funded 14 at the rank it had on paper, 31 would've sat another year, and the gap between a $40-a-ton overlay fix and a reconstruction job doesn't stay a gap for long on that kind of cracking.

We caught it because Dale happened to remember what 14 looked like under his wheels, not what the sheet said. That's the uncomfortable truth about windshield surveys: they're only as good as the crew's memory and the day's weather and whatever got driven last. A good rater catches this stuff sometimes. A tired one on month three of a six-month cycle might not, and nobody finds out until the overlay crew shows up and the pavement underneath doesn't match the work order.

What that week called for was a second source that didn't get tired, didn't skip 31 for fourteen months, and scored both segments the same way on the same day. Something that reads the whole network at once instead of whatever a truck happened to cover that quarter, so a 1.2-mile bad patch doesn't get to borrow credibility from the nine good miles around it, and a low-traffic road doesn't disappear from the list just because it's inconvenient to drive.

That's more or less the gap Road Surface Condition is built to close: a once-a-year pass over the full network from VHR imagery, classifying cracking, raveling and patching by segment so the ranked list going into budget season reflects the roads, not the route the crew drove to get to them.

We fixed CR 14 and CR 31 that Tuesday, mostly by luck and Dale's memory. Not every district has a Dale, and even the ones that do shouldn't have to rely on him catching it before the board meeting.

If your resurfacing list has ever been built on whatever got driven last, it might be worth seeing what the whole network looks like at once.

Action requested

If your paving list is due before the next windshield survey can finish, tell us about the network.

Start a pilot

← All field notes