The construction world might have just gotten a serious upgrade. Today DroneDeploy unveiled a product called Progress AI, which it’s billing as “a groundbreaking solution that delivers automated progress tracking from both aerial and ground capture data, at unmatched speed and cost.”
Truly groundbreaking? Maybe that’s a bit of hyperbole. But as an ambitious project from the company that initially launched as a drone mapping startup? It’s something worth watching, particular in this era of AI everything.
With Progress AI, construction teams will be able to better understand their sites by way of using what’s called vision-language AI. That can then translate drone and 360-degree imagery into real-time progress reports.
What is Progress AI?
Here’s how Progress AI works. Then the user uploads either drone footage or 360 walkthroughs (yes, DroneDeploy has been increasingly dipping into on-the-ground data in a smart move that has led to massive user growth).
Then, Progress AI steps in. It can automatically detect installed work, measures progress and delivers percent-complete breakdowns by floor, location and trade. This all typically happens within minutes. For the user, that means no manual plan-marking or status chasing. It’s hands-free documentation powered by AI.
Progress AI is built on DroneDeploy’s proprietary dataset of billions of square feet of field data. It can identify construction elements, flag risks, and even answer questions in natural language like, “Where are we behind on drywall?”
Why this matters for the drone industry
For folks in the construction vertical of the drone industry, the answer is clear. But Progress AI has interesting implications for other verticals that rely on drone data. Progress AI is an example of how vision-based autonomy is reshaping job sites.
Drones are very good at capturing data. DroneDeploy has been a leader in capturing even more of it with its robust mapping and inspection tools used across agriculture, energy, infrastructure and commercial construction. Now with Progress AI, the move is away from merely collecting data and toward interpreting it with context, nuance and speed.
By integrating 360 camera data with aerial drone imagery, DroneDeploy positions itself at the intersection of air and ground robotics. From there, it taps into what has become one of the drone industry’s bigger challenges: how to turn vast visual datasets into immediate, actionable information.
The rise of vision-language AI in drones
Progress AI is powered by vision-language models (VLMs). This is essentially of the next evolution beyond traditional large language models. While LLMs like ChatGPT understand and generate text, VLMs go one step further: they see.
That means AI doesn’t just analyze metadata or annotations — it interprets raw visual conditions in context. A VLM type of AI can understand that a duct is installed, that drywall is missing and that sequencing is off. It’s computer vision paired with linguistic reasoning, trained at scale using DroneDeploy’s dataset of real-world jobsite imagery.
In nerd speak, this is indicative of a move toward semantic perception. It’s where drones don’t just capture the world, they explain it.
Early access and what’s next
Progress AI is now available in early access and will officially roll out to all DroneDeploy Aerial and Ground customers in October 2025. For now, construction teams interested in joining the pilot can sign up at DroneDeploy’s website.
The post DroneDeploy’s Progress AI could mark a big moment for drone data appeared first on The Drone Girl.