When they fall from the sky.
It’s just a matter of time until someone is killed or injured by one. And I have no doubt that some are going to start using them as personal weapons.
5 thoughts on “Drones”
Comments are closed.
When they fall from the sky.
It’s just a matter of time until someone is killed or injured by one. And I have no doubt that some are going to start using them as personal weapons.
Comments are closed.
I wonder how hard it’d be to fly a drone into the air intake of jet engine on a passenger plane.
Damned hard – small target, huge closing velocity. This is a very tough problem for a homing missile with an active seeker; no way this is happening with a very low bandwidth control loop.
Meh, not an impressive article. At least they discussed the hideous mishap rate of the MQ-1 Predator though. The big problem with GA’s designs is the lack of auto-takeoff and auto-land; doing high-bandwidth tasks through a link with crappy visual feedback only is a recipe for disaster. Also worth noting that the Navy Global Hawk that crashed in Maryland was one of two prototypes that the Navy had been flying in test since 2006 or so, not a production bird.
The big technical problem with UAVs is see and avoid – it’s hard to envision integrating them robustly into, say, VFR takeoff and landing ops without the ability for eyes out. Aside from that, there’s no reason why a well designed UAV should be less safe than a manned vehicle. Of course, there are not many well designed UAVs out there though…
The Army’s MQ-7 Shadow and MQ-1C Gray Eagle (think of it as a Predator on steroids) have an automatic landing capability using a system called TALS. Gray Eagle also has an automatic takeoff capability while Shadow is launched on a catapault. Gray Eagle is built by General Atomics like the Predator but incorporates many improvements including a more powerful diesel engine, the ability to carry up to 4 Hellfire missiles, and longer duration.
I’d like to see an analysis of the accident rate over the past few years. UAV technology has increased dramatically since 2001 and I’d like to see how the rate is trending over time. Just stating the number of accidents over a ten year period isn’t as useful as providing a breakout year by year. It’d also be useful to look at the different UAVs and how their rates trend. It’s reasonable to expect the accident rate is higher in the early years of service, declines over time, and perhaps increases as a function of vehicle age. Lumping in all UAVs provides little useful information.
One difference in the accident rate, at least for takeoffs and landings, might be the operating philosophy of the different services. Army UAVs are normally controlled by enlisted people. The Air Force uses officers as pilots and they’re more likely to hand-fly the vehicle.
On a Northrop Grumman UAV, there is no way the pilot can hand-fly a takeoff or landing; the pilot has outer loop control only. Again, it all boils down to the near impossibility of doing high quality, high bandwidth inner loop control near the ground with a video camera for feedback. To my knowledge, no RQ-4 Global Hawk or MQ-8 Fire Scout has ever had a takeoff or landing accident that can be attributed to the auto takeoff or auto land algorithms.