Hacker News Reader

Comment by bencollier49 | Hacker News Reader

It feels like they're paying ludicrous money for these. Can someone in the know explain what differentiates these from shop-bought or scratch-built drones?

As per this article below, the Nano 1A UAV Quadcopter Bug drone sells for £4,500 (a tenth of the price of another model, the Black Hornet Nano):

https://www.independent.co.uk/news/uk/home-news/drone-nano-b...

I appreciate that this is probably a variation on the old canard that goes "I could have written that government software in a weekend", but it does seem pricey.

View on HN

There's maybe $300 of off-the-shelf components (at retail prices) in the pictured air unit. It's the R&D time. The software dev and testing of the the whole system would be super expensive. Also there's no mention of their ground station design, how is it controlled by the operator, there's a lot of money and research that has to go into that half of the problem. Soldiers wont put up with the operational peculiarities that hobbyist will. Drones are incredibly fragile because of the weight limitations. 40 minutes of flight time is amazing, probably 2x the best time a hobbyist could build with off-the-shelf parts at that size class.

It's a toothpick drone with two 18650 batteries on it, I doubt it costs more than $100. There's video above of a hobbyist doing the same thing. I don't know how autonomous this one is, but most quadcopters nowadays run STM32F4 processors, so, as you say, the software would be what's expensive.

From the article:

"It boasts a stealthy low visual profile and the ability to fly even in strong winds of more than 50mph. It was the only nano-UAV able to cope with the uncompromising weather during a recent Army Warfighting Experiment (AWE) event hosted by the Ministry of Defence’s Future Capability Group."

> It boasts a stealthy low visual profile

The one in the article's picture looks like a flying brick. A lot of money to be made in this field - there's a low barrier to entry in the market.

The problem as I see it is that many, many people have the skills to compete, but not the capital. And often are working so much that they don't have spare energy to invest every weekend for years to build an MVP.

I know some will say that sounds like making excuses, but the examples I've seen of entrepreneurs have been people who haven't actually had to work very hard at their day jobs. They aren't totally burned out at the end of the day.

I wish I knew a way to fix this situation. The economy could operate far more efficiently than it is, and that would be better for all of us.

Your comment should be made into article or at least maybe some question on economic stack-overflow. I too don't know solution yet. There are many people who have an idea, but many of those ideas will just fail because it's not sustainable or there are better ways already. The biggest problem is discovering those best ideas from noise. Some dedicated think tanks of specialists disseminating ideas would be best, but currently there is simply too much noise to filter.

"Army Warfighting Experiment (AWE) event hosted by the Ministry of Defence’s Future Capability Group."

I wonder who was allowed to attend... And how those were selected....

You need to separate R&D amortization from marginal price, otherwise you're deep in "ask a silly question, get a silly answer" territory.

Unless you're DJI the volume isn't high enough to just look at marginal price.

It's probably because of the low volume. They've only sold 30 so far, and the R&D costs need to be recovered.

If the selling point is stability in a storm, then the differentiation is in software, which is interesting. You'd expect the price of things like this to tumble.

One could imagine great dark clouds of these things in battle over perhaps Eastern Europe or the South China Sea in ten years' time.

And roving cities assigning tickets for illegal parking and all manner of other infractions.

Shop bought drones don't have arbitrary, often nonsense requirements. This one advertises that it'll work in 50MPH winds. How common are soldiers dealing with combat in 50MPH winds? It's likely there are other, similarly arbitrary and nonsensical demands that were placed on these.

Arguably one of the main differences between defence and consumer tech is dealing with the 'oh come on how often...'

You don't want to be caught needing your X or calling off operation Y because conditions are unfavourable.

It's kind of Pareto principle I suppose - they'll pay (say) 80% for that last rarer 20% of operational feasibility. (Likely more like 95-5, but I mention 80-20 as it's basically an 'aka' for Pareto.)

If you're fighting an opponent whose drones can only handle 40 mph winds, you might try to do as much combat at 50 mph as possible.

How many actual historical battles or skirmishes have been fought in gale force winds?

Operating nearby to helicopters may be feasible

https://www.smbc-comics.com/comic/rise-of-the-machines

"If you look at historical data, the vast majority of battle-winners used pre-modern weaponry"

"Thanks to the machine-learning algorithms, the robot apocalypse was short-lived"

Those wars are over. You have to be ready for the next war.

My guess it is built with military grade components which are more expensive and has got embedded emphasis on reliability. That probably includes custom molds that can be insanely expensive for small runs. I can imagine the firmware must conform to a higher standards than consumer grade firmware.

[x] Doubt

I've built dozens of multirotors and you can clearly see AMAXinno motors on the thing, (https://uavtek.co.uk/uav-fleet/bug) plus the big bulging 18650s. 99.9% chance it's using a COTS multirotor firmware or some variant of Ardupilot, and the camera looks like an old-school analog FPV camera. From the specs sounds like they have it running on the 5.8Ghz ISM band, so the thing will fall out of the sky the instant it encounter interference from some poorly configured consumer electronics.

I don't know for sure what's inside this drone, but it's probably because they had to pay for developing and building a custom small drone. I suspect DJI or Autel can build better drones at a far lower price, but the UK military probably does not want to rely on a Chinese company for their drones.

Maybe they've could have turned to Parrot, which is French? Or maybe even that would not be desirable, especially after Brexit.

I looked at the cost, 45k GBP each and thought "that seems like a steal compared to US Defense prices".

Reply on news.ycombinator.com