The Worst of Drones is Yet to Come...
Nine years ago I predicted how the rise of the drone will change war, politics, and society, when I gave the keynote talk to the Lockheed Martin executive lunch entitled, UAV: A Race to the Bottom. Yet, I am just baffled as to how miserably slow and low-tech the implementation has been to date, indicating at least to me a lack of vision and true understanding of what the drone is, and what it is not. Drones will in time destabilize vast political and economic structures – they will democratize conventional warfare, they will change narco and human trafficking, and will allow for “at whim” assassinations. What they are not is a way to get your package from Amazon or a Pizza. The press and media have gotten the message all wrong when they thought the drone would be a carrier of “hope” I see them as bringers of death and instability. So far as I can tell I am winning this argument. I have updated my predictive talk of 2015 with relevant citations from today that validate the arguments made back then.
My talk started with a basic truth: the last fighter pilot has already been born. Training a fighter pilot takes years and approximately $5.6 to 10.9 million dollars (Michael G. Mattock, Beth J. Asch, James Hosek, Michael Boito, 2019)). It is an expensive and time-consuming endeavor. Just look at how long it has taken to train Ukrainian pilots on Western aircraft; they were already fighter pilots when they entered training. In contrast, you can train a drone pilot in a day because the hard work of piloting is built into the drone and stored in its programming and hardware. An 8-year-old can fly a drone as well as any adult. This means that the drone can accomplish tasks that previously required great skill and training—such as navigating from one place to another with accuracy—and do it for $100 in parts rather than $5.6 million in training. The “issue” then is creating enough capacity to produce drones in massive quantities for the cheapest price – a race to the bottom (Skove, S, 2024).
When I spoke at LM, I was in the process of pitching a project for funding—to build an automated drone with a 1000 km range that can carry 1000 lbs for less than $10,000 per unit cost. The business model was based on the success of the V1 designed by the Nazis and used very successfully against the UK in World War II. The Germans were really the first “early adopters” of the drone as a bomb although the history of unmanned aircraft goes back to WW1 and earlier (if curious read the dissertation by BISHANE A. WHITMORE, 2016). The V1 wasn’t the most effective weapon but it was definitely the first mass-produced drone (Jeff Davis, Intergalactic).
My plan was to utilize an unused Honda factory in South America to mass produce the drone air-frames on an automotive scale – 100,000 a year. At that time, the power plant was a very cheap 150hp 2-stroke, and the airframe was based on the proven Long-EZ. The market was really remote delivery of supplies to mining, oil, relief supplies, etc – because people often fail to factor in the cost of operating not only includes machines and people, but security and complex logistics and bad weather. If the airframe is cheap all this doesn’t matter much.
The core concept of the talk was that we had started a global “race to the bottom,” and whoever can make drones faster, cheaper, NOT BETTER, is going to win the next arms or industrial race. The US military has always been built on the idea of 99.9+% reliability, I mean you want to protect your expensive pilot after all, whereas our design only needed to be 97% reliable. If you used it for carrying freight and supplies to remote areas writing off 3% is not a big deal. UPS insurance premiums are after all more than 3% and we pay that. Furthermore, although the plane could fly multiple missions it only needed to fly 1x and one-way to be economical. When it got to its destination the motor would shut off and a ballistic parachute deploy and it would have a soft crash landing. Recycle the airframe or not, it didn’t matter economically. I am still waiting for this to happen – it hasn’t.
Fast forward to today, the drone has become, according to the Ukrainian military, “the most important” hardware on the battlefield (Ukrainian Ministry of Defense, 2023). For example, in just two years, we transitioned from a $249,700 man-portable line-of-sight anti-tank missile to $500 over-the-horizon drones that are destroying $5 million tanks. The era of expensive equipment is over (Tomas Milasauskas, Liudvikas Ja?kūnas, 2024), and the 17,500+ armored vehicles lost in Ukraine should make that obvious. For my money, the bravest person on earth is now the driver of a T-80 tank who has to live in fear of a vape-smoking 19-year-old 10 km away, wearing a VR headset and listening to speed metal (Raj M. Shah and Christopher M. Kirchhoff , 2024).
Yet here we are still buying $82m F-35 fighters. I mean we need those, right? Play economist general for a second which would you rather have on the battlefield a single $82m plane of 8,200 flying bombs with a 1000km range? The F-35 has its place but the economics aren’t there for most of the nations in the world, and it still has to land, refuel, reload, and so on and those locations are vulnerable even if the plane might be able to survive in the air. It is, as is apparent in Ukraine, a race to the bottom that is working, not a race to quality.
However, as I mentioned at the beginning we are still far from where I thought we would be in the Droneverse. Nine years ago I pointed out that we already had terrain-following optical tech that was nearly a decade old and could run on an ARM processor. In practical terms, this means that drones should be flying completely autonomously using un-jammable optical at between 20 to 50m off the deck making them all but impossible to pick off rather than the 100-200m most seem to be operating at. Yet this hasn’t happened.
领英推荐
Just like their multi-million dollar expensive cruise missile brothers, a $200 cell phone can run good accurate maps and optical flow will allow you to fly by landmarks rather than GPS, making them mostly immune to electronic warfare. Yet this too hasn’t happened at scale or on the cheap but it is old-proven tech.
Also in my presentation, I showed that back in 2015 people were already flying armed drones capable of shooting. Pair that with thermal and facial recognition (again built into most smartphones) and you can have the drone do the dirty work of taking head shots at people on the ground. The tech is not “new,” it's old tech repurposed. In Ukraine, they are headed in that direction but are they there yet? No, again I am surprised this hasn’t happened.
Why is this development cycle so slow? I can’t really say. In my 30 years of experience running hardware products, I often see that engineers do what they know, rather than what can be done. Engineers are good at designing to directions, and some of the best inventors tend to be people who don’t have engineering degrees. Imagination often trumps education, science fiction readers might have better forward-leaning ideas than those who read engineering journals, and the drone is not an aircraft it is an electronic system that just happens to fly. In any case, what is obvious is that what is possible and proven still hasn’t been built and deployed and there has been plenty of time. It's baffling.
I will write about the societal upheaval to come in my next post. Until then…
References