The Rotomill Ate My Control Points

So I roll up to a site this morning. 16-month project, been flying it since February. I've got my control network dialed in — semi-permanent targets, paint over PK nails, spread across the site like they're supposed to be. Not clustered in one corner like a bunch of sheep. Five on the curbs, three on the asphalt edges, a couple on sidewalk corners. The kind of setup where you show up, set up the base, and you're flying in 20 minutes because everything's already marked.

Two weeks since the last flight. No big deal. It's what the Superintendent wanted.

I pull up to the east boundary and the street is gone. Not gone like "they patched a pothole." Gone like rotomilled down to the base course. Three of my points were on that asphalt edge. Three PK nails I drove in 6 months ago, painted fluorescent orange, logged in my control sheet with photos and descriptions.

Gone. Never to be used again.

But the gutter pans? Still there. Concrete doesn't mill. I had targets painted right in the flow line — bright white targets that had been repainted last visit on the gutter pan concrete. Those three survived. Those are the ones you bet on.

Nothing Is Permanent on a Construction Site

Here's the deal. When you set control for a long-term project, you're playing a game of predicting what the contractor won't destroy. You look at the plans. You talk to the super. You pick spots that seem safe — permanent structures, established curbs, things that were already there before you showed up.

And then the milling schedule comes through and none of that matters.

The curb you painted last month? Getting demo'd next Tuesday for a storm drain tie-in. The sidewalk corner you've been using as your checkpoint? They're widening the intersection in Phase 3. The asphalt edge with three of your best points? Mill-and-overlay, baby. Scheduled six months ago. The superintendent knew. The project manager knew. The paving sub definitely knew.

Nobody thought to tell the drone guy.

And honestly? That's not their job. It's mine. I'm the one who needs the control network. I'm the one who should be asking.

Why Gutter Pans Survive

Here's a trick I learned the hard way: paint your targets in the gutter pan.

The gutter pan is the concrete channel between the curb and the asphalt — the flow line where water runs. It's part of the curb-and-gutter structure. When the street gets rotomilled, they grind the asphalt down to the base course. The gutter pan? Untouched. It's concrete. The milling machine rolls right past it.

Same goes for curb demolition. When they tear out a curb section for a storm drain tie-in, they're taking the vertical face and maybe the top. The gutter pan at the bottom usually stays. It's below the work zone. It's not in the way.

I had three targets painted in the gutter pan on this site. Bright white, right in the flow line. Good contrast from 200 feet up. The asphalt points next to them got eaten. The gutter pan points? Still there. Still visible. Still usable.

Now, if the whole street section is getting rebuilt — full-depth reclamation, new curb-and-gutter, the works — then nothing survives. But if that's happening, you've got bigger problems than control points. You're re-establishing the entire network anyway.

For everything else — mill-and-overlay, spot curb repair, utility cuts — the gutter pan is your safest bet. Paint a cross. Make it bright. Log it with a photo. It'll be there when you come back.

Concrete doesn't mill. Remember that.

What I Should Have Done

Jeff Musser over at skybornedronesolutions.com runs 10 to 12 permanent GCPs on his solar projects — 350 acres, multi-year contracts. He doesn't guess where the safe spots are. He talks to the construction manager before he drives a single nail. He asks: what's getting torn up? What's on the paving schedule? Where are the underground utilities going in? Then he picks his spots.

I know this. I've heard him say it on the DMA calls. I just got comfortable. Two weeks off and I forgot the first rule of construction-site control: the site changes whether you're there or not.

The Recovery

So what do you do when three points vanish?

You re-establish. You don't process the flight with a degraded network and hope the software figures it out. Pix4D will tell you you're accurate to a quarter inch. Voodoo blackmagic bullshit. You don't trust the software's own report card.

I shot three new points. Drove fresh nails into spots I verified with the super — not on the milling schedule, not in the demo zone, not anywhere near the Phase 3 intersection work. Took an extra 45 minutes. Ate the time.

I always eat that extra time. GCPs whether the client pays for them or not. Because the workflow is perishable. You stop shooting checkpoints, you get rusty. You get rusty, your ortho drifts. Your ortho drifts, somebody's paying for it — and that somebody is eventually you, when the client stops calling.

On my Holly Hills job I run 19 GCPs. Nineteen. RMS accuracy comes in around 0.05 feet — five hundredths. That's well under the 0.1 ft target. Longest baseline under a thousand feet. Nearby GEONET station about two kilometers away for reference. Without those GCPs, the alignment drift is visible and unacceptable. The engineer would catch it. The PM would catch it. And I'd be re-flying on a Saturday.

Bottom line money.

The Workflow That Survives Rotomills

Here's what I do now. This isn't theory. This is what kept the other 16 points alive today while three got eaten.

1. Cook your base first. Set up the base station and let it collect. While it's cooking, you're walking the site. That's not dead time — that's recon time.

2. Talk to the super every visit. Not just at project kickoff. Every time you show up. "Anything change since last flight? Anything getting torn up? Anything getting paved?" They won't volunteer it. You have to ask. But they'll tell you if you ask.

3. Drive nails where they can't reach — and paint where they can't mill. Not on asphalt that's scheduled for mill-and-overlay. Not on curbs in the demo zone. Find the spots that survive the entire project lifecycle — permanent structures, monuments, things that were there before the project and will be there after. And put painted targets in the gutter pan. Concrete doesn't mill. Bright crosses in the flow line. They'll outlast the asphalt every time.

4. Document everything. Photos of every point. Descriptions. A control sheet you can hand to someone else and they'd find your points without you. Because someday you might not be the one flying the site, and the next guy needs to know what's what. Always look out for the next guy, it might be you.

5. Shoot GCPs every time. Even if the client didn't pay for them. Even if you're "just doing a quick progress flight." The workflow is perishable. Embed it or lose it.

The Bigger Point

Control networks aren't set-and-forget. They're living things on a living site. The rotomill didn't eat my points because I was careless — it ate them because construction happened, and construction always happens.

The difference between a hobbyist and a professional isn't whether things go wrong. Things always go wrong. The difference is whether you have a recovery workflow or you're standing there staring at milled asphalt hoping the software will save you.

It won't.

Stop flying without checkpoints. That's it. That's the whole post. Go set your targets.