flying three miles high, catastrophic collision. ( NO COLLISION OCCURRED )dc

[gnirtS]

The chances of looking in the right place at a time and seeing an object is one thing but also in that time frame to actually be able to accurately identify along with the size of the object and therefore distance is unlikely. Having no frame of reference makes these judgements very hard, even for much longer duration sightings.

I'd still say if you're flying a plane, your altitude is currently measured in Flight Levels and you see something fleetingly outside the last thing you should assume the object is, is a drone.
Unfortunately the AirProx board seems to always assume its a drone with absolutely no investigation or questions asked.

Theres a lack of awareness amongst pilots in general as to what drones are and what they can do. Most have never seen or used one and are totally unaware of the performance (or lack of) and their standard envelope. They get taken into the media hype, furthered by the unquestioning airprox and are more likely to blame a drone next time they see something unidentified.

I'm not saying there aren't problems - the report above contains several low altitude very plausible accounts as we all know there are complete idiots around operating drones but to blame EVERYTHING on one despite the parameters making it physically impossible is just crazy and doesn't solve anything.

 

[sar104]

The chances of looking in the right place at a time and seeing an object is one thing but also in that time frame to actually be able to accurately identify along with the size of the object and therefore distance is unlikely. Having no frame of reference makes these judgements very hard, even for much longer duration sightings.

For an unknown object of unknown size, agreed. But a Phantom, for example, is extremely recognizable. I guarantee that if you caught even a glimpse of one within 50 - 100 ft from an aircraft cockpit that you would have no trouble identifying it.

 

[gnirtS]

Because you know what a phantom looks like. And thats the issue - most pilots have never seen one or physically been around one to get an idea of physical size.

Identification is much easier when your brain is already accustomed to identifying such objects based on experience.

 

[sar104]

Because you know what a phantom looks like. And thats the issue - most pilots have never seen one or physically been around one to get an idea of physical size.

Identification is much easier when your brain is already accustomed to identifying such objects based on experience.

Yes - but claiming that most pilots don't know what a Phantom looks like is a bit of a stretch. Anyone with an interest in aviation is at least going to have seen photos of them - posted with most news stories about near misses.

 

[gnirtS]

Photos arent really any use - they don't scale it at all. Theres no real frame of reference involved there.

There's an argument for giving pilots familiarisation sessions with them at least.

Or in the UK at least, making the airprox board actually use some!

 

[pftarch]

You probably should read the actual ASSURE UAS collision studies on structural component damage and engine ingestion, rather than news articles about them. The results are not at all reassuring (no pun intended).

I'm not going to say that a sUAS collision can't possibly damage a commercial aircraft.

However, my concerns with the collision study that you reference are three fold:

1. I am always suspect of computer simulations. (I don't feel that this is always a deliberate malicious effort to show one side or another, and I am reasonably sure that the people doing the simulations were objective, but, it is EXTREMELY difficult to accurately predict real word physical outcomes when there are literally millions of variables at play in a fluid situation.) They are a reasonable STARTING point for analysis, but I don't believe that they constitute a FINAL definition of the situation. If DJI wants to keep selling drones in the US, they REALLY need to pony up and pay to launch some REAL Phantoms and Mavics at REAL jets and into REAL jet engines.

2. Why is it that when these reports are generated the weights are never understated for the sUAS (and for this one I feel that they ARE understated for the bird strike). The simulated UAS weight is four pounds in many of the examples and the simulated sUAS appears to be a Phantom. Four pounds is nearly 33% heavier than an actual Phantom 4, over 40% heavier than a Phantom 3 and over ONE HUNDRED percent heavier than a Mavic Pro. (This is sad as they pointed out the weights of the Phantom 3.) When dealing with a mass times velocity scenario, overstating mass will have a significant impact on final results, and in this case, often the results exaggerate the extent of damage caused by some of the most common personal sUAS's in the United States. The report has, I feel, on several occasions overstated the density of the sUAS. A jet hitting a 2 pound rock is going to receive more damage than if it hits a two pound down comforter. They also compare bird strikes using a four pound bird. Geese strikes are common in the US and the median weight of a goose is around 10.5 lbs. If they used a goose for the bird strike simulation they have understated the bird weight by 150 percent. The report says "sUAS collisions caused greater structural damage than bird strikes for equivalent impact energy levels (i.e. equal mass and impact velocity)" which is theoretically "accurate" for what it is, but I feel is misleading. The reality is severe bird strikes involves birds SIGNIFICANTLY larger than 4 pounds. If you are going to compare a common sUAS vs. a common bird in a mass times velocity calculation, PLEASE use the common weights of these items.

3. The report fails to address probably the MOST important research question, which is, "What is the LIKELIHOOD of a sUAS to commercial aircraft collision?" I'm not saying that it can't or won't happen, but, they refer to a report done in 2011 that involved Predator drones operating at SIGNIFICANTLY greater altitudes than we are legally allowed to operate at. Although it's not "impossible" for my Mavic to hit the vertical tail of a commercial jet in flight, I'm curious as to how statistically likely it is. It is possible for a stray meteor strike on a commercial jet in flight. Both are "possible" but, I feel, highly unlikely.

 

[sar104]

I'm not going to say that a sUAS collision can't possibly damage a commercial aircraft.

However, my concerns with the collision study that you reference are three fold:

1. I am always suspect of computer simulations. (I don't feel that this is always a deliberate malicious effort to show one side or another, and I am reasonably sure that the people doing the simulations were objective, but, it is EXTREMELY difficult to accurately predict real word physical outcomes when there are literally millions of variables at play in a fluid situation.) They are a reasonable STARTING point for analysis, but I don't believe that they constitute a FINAL definition of the situation. If DJI wants to keep selling drones in the US, they REALLY need to pony up and pay to launch some REAL Phantoms and Mavics at REAL jets and into REAL jet engines.

I was not commenting on the methods - I was responding to a poster who quoted a news article and implied that this study concluded that drone collisions were not likely to cause significant damage. That's not its conclusion.

Direct testing is always preferable but when the parameter space is very large, which it is in this situation, numerical modeling is a well-established approach and perfectly reasonable. It's also been used extensively for bird strike damage assessment. All the reasons that you state to distrust modeling are, in fact, even stronger reasons to distrust experiments. Sufficient experimentation to explore that parameter space with statistical confidence is going to take a long time and be very expensive. DJI is very unlikely to fund such tests, and there is no obligation on them to do so.

2. Why is it that when these reports are generated the weights are never understated for the sUAS (and for this one I feel that they ARE understated for the bird strike). The simulated UAS weight is four pounds in many of the examples and the simulated sUAS appears to be a Phantom. Four pounds is nearly 33% heavier than an actual Phantom 4, over 40% heavier than a Phantom 3 and over ONE HUNDRED percent heavier than a Mavic Pro. (This is sad as they pointed out the weights of the Phantom 3.) When dealing with a mass times velocity scenario, overstating mass will have a significant impact on final results, and in this case, often the results exaggerate the extent of damage caused by some of the most common personal sUAS's in the United States. The report has, I feel, on several occasions overstated the density of the sUAS. A jet hitting a 2 pound rock is going to receive more damage than if it hits a two pound down comforter. They also compare bird strikes using a four pound bird. Geese strikes are common in the US and the median weight of a goose is around 10.5 lbs. If they used a goose for the bird strike simulation they have understated the bird weight by 150 percent. The report says "sUAS collisions caused greater structural damage than bird strikes for equivalent impact energy levels (i.e. equal mass and impact velocity)" which is theoretically "accurate" for what it is, but I feel is misleading. The reality is severe bird strikes involves birds SIGNIFICANTLY larger than 4 pounds. If you are going to compare a common sUAS vs. a common bird in a mass times velocity calculation, PLEASE use the common weights of these items.

I don't understand these arguments at all. Drones come in all shapes and sizes, and there is nothing unique about the Phantom. Since the risk clearly does go up with mass it's not surprising that they choose a conservative mass value. Yes - a 4 lb drone will produce more damage than a 2 lb drone. Are you hoping that while a4 lb drone causes severe damage to control surfaces and engine parts, a 2 lb drone collision causes insignificant damage and isn't a problem? If so you are going to be disappointed.

As for the comparison with bird strikes, of course they choose equivalent masses because that's the question being addressed - the effect of the mechanical properties of the two objects. When doing that you don't change other independent variables at the same time.

3. The report fails to address probably the MOST important research question, which is, "What is the LIKELIHOOD of a sUAS to commercial aircraft collision?" I'm not saying that it can't or won't happen, but, they refer to a report done in 2011 that involved Predator drones operating at SIGNIFICANTLY greater altitudes than we are legally allowed to operate at. Although it's not "impossible" for my Mavic to hit the vertical tail of a commercial jet in flight, I'm curious as to how statistically likely it is. It is possible for a stray meteor strike on a commercial jet in flight. Both are "possible" but, I feel, highly unlikely.

No - that's relevant and important to risk determination, but completely outside the scope of this study. Risk convolves probability and consequence, and this is a study of consequence. As a comparison example, when impact testing is performed and reported on vehicles nothing is included about the probability of the collisions occurring. This study was strictly about the effects of such collisions.

 

[Air Beetle]

No, without reporting the probability, the consequence study creates hysteria. I agree 100% with pftarch. When we see vehicle impact tests we have a wealth of experience and we can easily determine the probability based on our experience as drivers. A drone to aircraft collision is overblown. I own an aircraft and fly drones. I try to educate other airplane pilots that the probability of a drone strike is much LOWER than a bird strike. There are 400 billion birds in the world.

 

[gnirtS]

Yes but sensible risk management involves reducing that risk as much as possible BEFORE an incident.
Birds you have limited control over (yes you can have scarers etc but not full control). Drones you CAN exercise a lot more control over and therefore reduce the risk of a collision.
We know people are flying drones illegally high, illegally close to airports and everything else. There are legit, verified near misses. We know an impact can cause damage. So far we only know of one confirmed impact.
So the potential risk is there and should be addressed.

 

[sar104]

No, without reporting the probability, the consequence study creates hysteria. I agree 100% with pftarch. When we see vehicle impact tests we have a wealth of experience and we can easily determine the probability based on our experience as drivers. A drone to aircraft collision is overblown. I own an aircraft and fly drones. I try to educate other airplane pilots that the probability of a drone strike is much LOWER than a bird strike. There are 400 billion birds in the world.

I don't see any hysteria - where is that being reported? In any case, those are quite different studies - a group doing numerical simulation of collision is not going to have expertise in aviation traffic conflict assessment. Someone else needs to do that.

 

[ac0j]

There sure does seem to be a lot of news about drone /aircraft near misses and how dangerous that is. But there is no news proving that it is that dangerous because it has not happened yet. A helicopter got a couple dents once.
Could be considered hysteria. Hyping it up when nothing bad has ever happened.

 

[gnirtS]

"couple of dents". Tens of thousands of dollars worth of damage and a damaged gearbox which IS a flight critical component. Or are you claiming expensive damage isn't important?

You can throw rocks at moving cars, it'll put a "couple dents" but wont cause them to crash - does that mean its OK to throw rocks at cars?

 

[sar104]

There sure does seem to be a lot of news about drone /aircraft near misses and how dangerous that is. But there is no news proving that it is that dangerous because it has not happened yet. A helicopter got a couple dents once.
Could be considered hysteria. Hyping it up when nothing bad has ever happened.

Near misses are required to be reported. Most of the hysteria that I've seen resides on this forum and others like it, and comprises frantic denials. Some of the reports are clearly problematic, but others are quite credible.

 

[AnzacJack]

the experts said that it couldn't happen

 

[cayman5522]

Monday 17 September 2018
According to the Doncaster Free Press.
A passenger plane with 240 people on board narrowly avoided a catastrophic collision with a drone as it took off from Doncaster’s Robin Hood Airport.


The drone, which was flying three miles high, was just 100 feet away from the Airbus A321 which was climbing at 380mph after taking off from the airport.

The incident happened at the highest ever altitude recorded for a near-miss between a drone and an aircraft in UK airspace.

Packed passenger plane nearly hits drone at Doncaster Airport

The same pilot states he saw E.T. and Predator up there....

 

[MrMavic]

Malarky I say. Just another weak kneed attempt to assassinate and degrade our sport. There's no such thing as a near miss. "Near Miss" would be a crash, right. When you nearly miss, that means you crashed into whatever. Let's just call it a miss for crying out loud.

 

[jsh102]

Out of curiosity, how do you get the MP to fly above 400 feet? Mine gets up to that elevation and holds. I've never really been in a position to go higher or really needed to, but how would it be done?

 

[Static8]

3 miles high? Wow! It's serious if true but calling 100 feet near? Or catastrophic? It just goes to show that it doesn't matter what laws you put in place, criminals don't care.

Wait, isn't this kinda convenient with the drone laws being discussed in the UK? Hmmm....

This is exactly what occured in Canada. There were several “near misses” reported by media and or government when new laws were being discussed or implemented...Classic BS ..the world governements on the most part, do not want consumer drones in our skies. I think its about control and they are afraid we will see something they are hiding.

 

[Lon Denard]

I agree with your comments about the issues with drones at the altitudes reported, but not about the theoretical visibility of a drone at aircraft speeds. 0.75 seconds is ample time to notice and identify an object moving with parallax against a distant background.

As a simple direct comparison, since speed and length scale linearly, consider a white, one-inch object flying past (or into) your car windshield at 50 mph. I'm pretty sure that you have easily noticed and identified insects as small as that - I certainly have.

Absolutely correct. I've spotted and id'd all sorts of small objects traveling over 150 mph on my bike. It has more to do with processing speed than visual acuity. I trained for target acquisition and identification, it's definitely something you can improve.

 

[4wd]

Out of curiosity, how do you get the MP to fly above 400 feet? Mine gets up to that elevation and holds. I've never really been in a position to go higher or really needed to, but how would it be done?

You can change it to 500 metres about 1600 feet in the settings.
There are ways to go higher but it takes time to go up and descend, no small drones are going 3 miles high, unless on a one way trip to oblivion.