For drone pilots who already map construction sites with drones and 360‑degree cameras, your imagery just got a lot more valuable. DroneDeploy has unveiled Progress AI, an agentic vision‑language model (VLM) that chews through aerial maps or 360 panoramas and spits out color‑coded progress reports — in about the time it takes to grab a coffee.
Most construction crews still lean on manual walk‑throughs, colored paper plans, and scattered field notes to work out what’s finished and what’s running late. High-end progress-tracking software exists, but price tags and setup headaches keep those tools locked to marquee jobs. Progress AI flips that script:
- Minutes, not days: As soon as you upload a drone map or 360 pano set, the AI calculates percent‑complete for every trade and floor. No aligning BIM models or hand‑drawing heat maps.
- Voice or chat queries: Need a quick answer? Ask, “How far along is Level 4 drywall?” and the system reads its own “photographic memory” to reply — whether the data is from yesterday or two years ago.
- Affordable scale: DroneDeploy says the add‑on costs only a fraction of existing solutions, so even mid‑sized remodels can justify the spend.
For Part 107 drone pilots who already collect imagery for clients, that speed and price combo could turn routine flights into a premium analytics service.
How the tech works
Progress AI isn’t just a large language model reading spec sheets. It’s a vision‑language model trained on “billions of square feet and millions of acres” of ground‑level and aerial visuals. Feed it a stitched orthomosaic or a 360 walk, and the model identifies installed work — framing, MEP, finishes — then cross‑checks location, sequence, and visible changes to flag what’s complete and what still needs attention.
During beta tests, superintendents used the tool to:
- Validate subcontractor pay applications in hours instead of days.
- Spot out‑of‑sequence trade stacking before it triggered schedule slippage.
- Catch issues like a mis‑hung door frame purely from the AI’s room‑by‑room heat‑map.
Wharton‑Smith Construction, an early partner supporting 90 projects nationally, says the system lets its digital documentation team “check in on progress without setting foot on‑site.”
Here’s what the workflow looks like:
- Capture as usual: Fly the site with a drone or walk it with a 360 camera. No special flight paths required.
- Upload to DroneDeploy: Once the pano set or processed map is in the cloud, Progress AI kicks in automatically.
- Get answers: View the auto‑generated dashboard or just ask a question via voice or text.
Because the tool doesn’t demand BIM files or complex schedule integrations, field crews don’t have to overhaul their current routines.
Here’s who benefits from DroneDeploy Progress AI:
- Superintendents and PMs: Free up hours spent coloring drawings and compiling daily logs.
- Trade partners: Resolve sequencing conflicts faster with objective visuals.
- Owners and execs: Verify schedule health across entire portfolios without hopping on a plane.
- Drone service providers: Package instant progress analytics alongside imagery to upsell clients.
In effect, Progress AI acts like an extra set of superintendent eyes, only this “new hire” never sleeps and remembers every capture.
Basically, Progress AI compresses what used to be days of manual progress tracking into a coffee‑break‑length task, all powered by the imagery many drone pilots already collect. If the rollout lives up to beta feedback, expect construction clients to start asking for more than just pretty orthomosaics: they’ll want the instant, room‑by‑room answers that come with them.
More: Gaussian Splatting magic: DJI Terra’s biggest leap since launch
FTC: We use income earning auto affiliate links. More.
Comments