PPRuNe Forums - View Single Post - Automation vs Seat-of-the-pants-flying talking as devil's advocate - so no abuse plea
Old 27th Jul 2013, 16:04
  #52 (permalink)  
deptrai
 
Join Date: Nov 2009
Location: flying by night
Posts: 500
Likes: 0
Received 0 Likes on 0 Posts
FedEx-founder Fred Smith posted this in 2009:

Unmanned cargo freighters have lots of advantages for FedEx: safer, cheaper, and much larger capacity. The ideal form is the 'blended wing.' That design doesn't make a clear a distinction between wings and body, so almost all the interior of both can be used for cargo. The result is that the price premium for air over sea would fall from 10x to 2X (with all the speed advantages of air).
and

the key thing is having NO people on board, not even as backup. A single person in the craft requires a completely different design, along with radically different economics and logistics. The efficiencies come with 100% robotic operation.
What he says mostly makes sense to me (my biggest doubt at this point is that he claims it will be "safer", that remains to be proven, but no human pilots on board may well be safer in terms of loss of human lives, if the things can manage to avoid crashing into humans on the ground and in other airplanes). It's also interesting to note that the unmanned X-47B which just completed first carrier landings is a blended wing design.

In response to VinRouge: following Smith's line of thinking, I'm not sure if we will see single pilot operated transport category cargo aircraft first. It's not a necessary intermediate step to what he wants. A more likely evolution, as I see it, would be to let Boeing gather experience with the Phantom Ray, maybe they'll even win the contract for the US Navy UCLASS and then, much later, when the military technology matures, start to work with them to look at the feasibility of larger civil cargo UAVs.

You also mentioned that
UAVs [...] by and large still have a human operator to land them (Predator is landed on site and is not autoland)
This may be true right now, but it will change. Carrier-based UAVs use task based automation, including autoland, mainly because there is no current datalink technology which is fast and powerful enough to allow something as precise as a carrier landing to be remote controlled, and autoland on the other hand is known technology.

But for civil aviation this is not even at the stage of long term strategic planning, but merely a visionary concept. Four years later, we're not really much closer to Fred Smith's aspirational idea. It will no doubt happen, that's not the question. The question is when? There I agree with previous posters, it will take at least two decades, most likely longer, with a caveat that it's hard to predict anything, particularly the future.

Just applying military experience to design civil UAVs won't be enough, there's many civil aviation specific issues that will need to be addressed. Military UAVs mostly operate in a very different kind of airspace (think of separation and control). Civilian (partly) autonomous/automated UAV would need much better sense-and-avoid technology, eg the ability to interpret video images, radar, as well as transponder/ADS-B data, to maintain separation. Also, the military owns and operates satellites and has huge network bandwidth at it's fingertips which can relay images to the ground with short latency. Civilian/Commercial UAV operators would have to buy bandwidth, and even compete for the spectrum with other uses if they want to t/x huge amounts of data to allow for remote control. And finally you also need FAA to certify all this new technology which could easily add 20 years alone...

Last edited by deptrai; 28th Jul 2013 at 10:32. Reason: damn tpyos
deptrai is offline