A look at the burgeoning drone software and AI industry, a focus for an estimated 65 US startups, often for military uses
Networked drones predicted to unleash host of new applications — In San Diego, the defence start-up Shield AI is designing software that will fly drones without any human involvement. Tweets: @thuttag and @carlbfrey Tweets: Thu Thu Aung / @thuttag : Drone innovation moves from hardware to self-flying software https://giftarticle.ft.com/... “[We are at] the tip of the iceberg in terms of capabilities that are going to be unlocked by software,” Adam Bry, Skydio's chief executive, said. Carl Benedikt Frey / @carlbfrey : According to the Federal Aviation Administration, the use of registered industrial drones surged from 12,000 in 2015 to nearly 500,000 by 2020. This “appears to be at an inflection point demonstrating powerful stages of growth”. https://www.ft.com/...
Context & Ripple Effects
The drone startup wave of the mid-2010s ended badly: after hundreds of millions in aggregate fundraising, many of those companies had closed or pivoted to services once the consumer hardware market matured. What survived changed shape — when the US blacklisted China's DJI, Skydio's $171M raise led by a16z at a $1B+ valuation marked the first significant US drone funding of that era, signaling that value was migrating from airframes to software.
This Financial Times piece captures where that migration has landed: an estimated 65 US drone software and AI startups, often serving military customers, with San Diego's Shield AI building systems that fly drones with no human involvement and Skydio's Adam Bry arguing the industry is only at the tip of the iceberg of what software unlocks. It matters because it reframes drones as an autonomy problem — and sets up a direct contest with China's state-backed push into its low-altitude economy.
First-order effects
- Shield AI and the other autonomy-focused startups are competing directly for US military software contracts, making defense procurement — not consumer sales — their primary revenue path.
- Skydio, fresh off its post-DJI-blacklist funding round, is positioned as the domestic alternative in a market where Chinese hardware dominance pushed US buyers toward homegrown suppliers.
Second-order effects
- China's low-altitude economy strategy answers the US software push with scale: leveraging its domination of drone production and R&D below 1,000 meters, forcing Western startups to differentiate on autonomy rather than cost.
- Suppliers and integrators around the FAA's registered fleet — up from 12,000 industrial drones in 2015 to nearly 500,000 by 2020 — become the customer base for retrofitting autonomy onto aircraft already flying, a larger addressable market than new airframe sales.
Third-order effects
- If autonomy software becomes the industry's center of gravity, drone manufacturing consolidates around a few platform owners whose flight code, not airframes, is the moat — repeating the pattern of the earlier hardware shakeout one layer up the stack.
- Military-grade autonomous flight spreading across dozens of startups points toward regulatory pressure on beyond-visual-line-of-sight rules and a bifurcated global market split along US-China supply lines.
The trend: Drone economics are shifting from hardware manufacturing to self-flying software, with US defense demand and China's low-altitude economy push turning autonomy into a geopolitical procurement battleground.