Flying over towns and cities?

[aesculus]

Just as a side note, they have tested the penny theory and there is no threat to human life as the velocity is minimal due to air resistance.

Yes, well we all live in the vacuum of our own opinions don't we? [emoji16]

 

[AyeYo]

I agree with all the safety considerations you mention, but there is way too much blasting people with the end of the world fire and brimstone when nothing has actually happened yet. I too dread the day that an innocent bystander is critically wounded from a hobbyists drone. But since that is not the case, more supportive recommendations and less holier than thou scolding seems more prudent.

I think the issue is that many of us recognize that when that day comes, the hobby is done. Too many people don't take that seriously enough.

 

[kneeL]

I hate to stick my nose where it doesn't belong - I love my Mavic and its awesome but yes it is "toy-grade" relative to the context in which the first person said it was toy grade (relative to commercial drones). Any expert pilot or commercial/industrial/gov't sUAS pilot would know exactly what they meant.

it's not a "toy" relative to the $100 to $250 range of quad-copter toys, correct. But that was not the context in which he made the comment.

So probably could just change the wording to "consumer-grade" and we can all be happy.

 

[jtc44]

I've been involved in aviation since a kid. Pilot, instructor. Navigation system software and firmware. Managing such product lines and selling them. (Military and commercial worldwide).

The systems and integration on a Mavic Pro for a "non aviation" product are outstanding and sophisticated and would have been unthinkable merely 10 years ago.

DJI could improve their bag of tricks IMO (GPS/IMU integration in particular) but they are head and shoulders ahead of everyone else.

The value in a product like the MP is incredible with its sole short marks going to its camera sensor.

In short whether it is a "toy" or a "tool" or a "professional" drone has little to do with the Mavic Pro. It has all to do with the person operating it.

I wholeheartedly agree and that's why I'll be purchasing one very soon. I won't, however, be flying over towns and cities as the thread states in its current iteration.

 

[aesculus]

I think the issue is that many of us recognize that when that day comes, the hobby is done. Too many people don't take that seriously enough.

Well, you've hit the nail right on the head there, haven't you? The bit of legislation that just came out up North should be a loud warning siren to us 'Mercans. I think the majority of the holier-than-thou posts scolding the people asking legitimate, practical questions about things like 400' AGL and nighttime flying, and VLOS, etc., are all motivated by the constant fear that public opinion will eventually turn firmly and permanently against the aerial hobbyist, which will lead to public outcry, which will lead to unnecessarily restrictive laws, which will lead me to find another hobby to spend money on.

I get it. I wish there wasn't so much interest in it either. I wish it was still a nerdy hobby that only the most dedicated would pursue, because you had to build whatever you flew, and you knew the aircraft inside and out and understood all risks involved. Back in those days, we welcomed noobs to the hobby. In my club, there was always a spirit of camaraderie and acceptance. We encouraged new hobbyists. We helped them solve problems the right way, by giving them knowledge.

But new technology has made owning and operating these aircraft easy on an unprecedented level. You literally don't need to know anything about aerodynamics or electronics, or batteries or whatever to get one of these things into the air and bring it back down in one piece. There are people pouring in to our little nerd niche in droves and it will only take one of them to **** it up for the rest of us.

So, to avoid that inevitability, we just preach fire and brimstone at the noobs, hoping that some of them will just decide it's not worth the investment. Drones break too easily, you can't fly them anywhere fun, you have to call the airport every time you take off, yada yada yada. That strategy will work on some. You'll keep a few out by simply being an AH to them, but not all. There's a dude out there who thinks it would be cool as **** to line up his drone into the path of a Cessna just to see the look on the pilot's face. That's the dude that will ruin it for the rest of us. And that dude doesn't browse the Mavic Pilots forum, so don't bother preaching at the noobs as though he's among them. If he cared enough to participate in the drone community, he would't be doing stupid stuff.

 

[bk_flyguy]

I don't think rationally pointing out the real dangers to people when a drone is flown over cityscapes is fire and brimstone, I haven't seen any posts here that come remotely close to that. And back when drones were a smaller niche hobby, I never heard of or saw anyone flying them around Manhattan island in between buildings. We just need to be careful and aware.

As soon as you enter an area with tons of metallic objects and line of sight obstructions, wireless signals, etc. and then fill it with a ton of people, your probability of having something go wrong multiplied with the probability of hurting someone if something goes wrong skyrockets.

 

[BorisTheSpider]

Seems to me that the Mavic has quite a bit of redundancy built in with many of its systems. For example, two IMUs, and two compasses, and the RTH fail-safe. What redundancies are you referring to that commercial drones have?

I won't talk about commercial drones because I know very little about that subject, but I can talk about manned aircraft, where one of the basic requirements for certification is that no single failure should be able to lead to a catastrophic outcome, ie. an uncontrolled crash.

That's absolutely not the case with even a basic quadcopter. Immediately I can think of the following components, any one of which failing will always lead to an uncontrollable crash: prop, prop mount, motor, motor wiring, ESC, frame/arm (reassuringly constructed from only the finest plastics available), CPU/FC, motherboard, battery, battery controller electronics, battery contacts, battery wiring and solder joints, battery retention clips.

In addition, it would be totally unacceptable in a manned aircraft to have a situation where the flight controls could just stop working, they can easily do this in a drone simply because of interference, jamming, a receiver failure, a transmitter failure, a software crash, a transmitter battery failure, flying the **** thing behind a hill/tree/building and probably more events that I can't think of right now.

We're still in "basic quad" territory with all these possibly catastrophic failures. When we add more complexity with GPS and so forth, it gets worse not better. What happens when, just like my phone or car satnav, a reflection or other error makes the GPS report an incorrect position (and don't tell me your satnav has never shown you on a parallel road a few meters or tens of meters away)? Now our "smart" drone isn't so smart as it uncontrollably flies itself 20 meters in a random direction trying to get where it thinks it should be whilst we watch someone's face/car/grandma getting bigger in the smartphone app we're using as a flight control system. Even ignoring GPS failure modes, GPS is barely fit for aviation applications when it's working properly, there's a reason we're still not routinely flying GPS approaches in airliners and don't tell me anyone checks RAIM before flying their drone.

Then there's the pilot who has no training, no license and no medical.

Having two compasses, or two IMUs, when both of them are just being used to compute a single position and attitude solution is not redundancy. In fact, it can at times be worse than having just one since there's no third source to "vote out" the bad source, as evidenced by the "compass disagree" failures in the mavic that people have been having where it can revert to atti mode and throw out the GPS due to a noisy back compass despite still having perfectly good data from two sources (GPS and front compass).

Redundancy is two entirely separate, not interconnected systems that can do the same thing, they must not be interconnected or providing inputs to a single point of failure further up the chain and they must not be able to **** with one another no matter what fails or how bad things get - there's absolutely no redundancy whatsoever in a mavic, categorically none at all.

RTH is not a fail-safe, it's a fail dangerous. When the controller disconnects you have a pilotless aircraft doing something that's at best less dangerous than the alternative, but can only be described as safe in the loosest possible sense of the term. I am by the way a big fan of the feature - it is a great idea and works well if properly configured and in good conditions, but if you want to see how dangerous it can be you only have to look at all the crashes reported on here a significant proportion of which happened during an autonomous RTH.

If all of this sounds over the top, allow me to invite you to purchase tickets on my new budget airliner which will be coming to market soon and bringing ticket costs down to unprecedented value pricing by using a combination of off-the-shelf software from smartphones, mass produced Chinese hardware that works as it's supposed to at least 99% of the time and unlicensed pilots who learned by trial and error.

Finally, as I said, I'm wholeheartedly in favour of reasonably unregulated drone flying - it's fun and safe provided people don't get carried away thinking these things are idiot-proof or reliable. There's always a big enough idiot to break the idiot proofing (hint: it's normally you) and Murphy's law ALWAYS applies.

 

[LeMur]

Mavic, Phantom, Inspire are expensive, smart, incredible "pieces of technology", but definitely TOYS. And very dangerous toys.

 

[Wombat55]

I won't talk about commercial drones because I know very little about that subject, but I can talk about manned aircraft, where one of the basic requirements for certification is that no single failure should be able to lead to a catastrophic outcome, ie. an uncontrolled crash.

That's absolutely not the case with even a basic quadcopter. Immediately I can think of the following components, any one of which failing will always lead to an uncontrollable crash: prop, prop mount, motor, motor wiring, ESC, CPU/FC, motherboard, battery, battery controller electronics, battery contacts, battery wiring and solder joints, battery retention clips.

Well, you haven't really defined what you mean by "single failure". But if you simply mean it literally, as you appear to be doing when your list such unlikely factors such as a sudden failure of "battery wiring" or "solder joints" or "battery retention clips", then let me point out that commercial aircraft are also vulnerable to "single failure" causing a crash. For example, if one of the wing suddenly fails, that's an unrecoverable "single failure" that can lead to a crash. Not a likely event (but, then again, neither is a sudden failure of a "battery contact"), but I do recall that there was a crash involving a USAF Thunderbird or Navy Blue Angel due to a wing failure as a result of an unnoticed stress crack. Want another example of a "single-point failure"? How about the failure of the horizontal stabilizer of Alaska Airlines 261? ( Alaska Airlines Flight 261 - Wikipedia ). One part failed (the horizontal stabilizer's jackscrew assembly) and the plane went down.

Now if you now want to argue that such events are unlikely and do not qualify as realistic "single-point failures", then I have to ask you the following: How many Mavics do you think have gone down due to failures in "battery wiring" or "solder joints" or "battery retention clips"?

 

[Wombat55]

Having two compasses, or two IMUs, when both of them are just being used to compute a single position and attitude solution is not redundancy. In fact, it can at times be worse than having just one since there's no third source to "vote out" the bad source, as evidenced by the "compass disagree" failures in the mavic that people have been having where it can revert to atti mode and throw out the GPS due to a noisy back compass.

Redundancy is two entirely separate, not interconnected systems that can do the same thing, they must not be interconnected or providing inputs to a single point of failure further up the chain and they must not be able to **** with one another no matter what fails or how bad things get - there's absolutely no redundancy whatsoever in a mavic, categorically none at all.

That's a ridiculous statement. You know very well that there is not complete redundancy of all flight controls and flight surfaces on commercial and military aircraft. For your definition of "redundancy" to be true, there would have to be two of EVERYTHING related to flight on every aircraft, including two separate and independent sets of horizontal and vertical stabilizers, and we all know that that isn't so.

As for the statement that having two compasses and two IMUs provides no redundancy because there is no third vote, that's an overly simplistic view. Certainly, greater redundancy is provided by having at least three independent systems because then a simple "majority vote" algorithm can be applied. "Majority vote" works in the case of either a "soft failure" of one of the units (i.e., fails in such a way that the unit is still operating but gives erroneous but possibly plausible outputs) or a "hard failure" of one of the units (i.e., fails completely and goes dead or outputs completely nonsensical data). Granted, if one has just two IMUs, it becomes difficult to impossible to detect a "soft failure" of one or more units, but a "hard failure" involving one of the units is easily detectable. There is redundancy. If you don't believe that, try telling me that you wouldn't prefer to have two GPS units on you rather than just one if you're wandering around alone and lost in the wilderness?

 

[AlanTheBeast]

I won't talk about commercial drones because I know very little about that subject, but I can talk about manned aircraft, where one of the basic requirements for certification is that no single failure should be able to lead to a catastrophic outcome, ie. an uncontrolled crash.

That's absolutely not the case with even a basic quadcopter. Immediately I can think of the following components, any one of which failing will always lead to an uncontrollable crash: prop, prop mount, motor, motor wiring, ESC, frame/arm (reassuringly constructed from only the finest plastics available), CPU/FC, motherboard, battery, battery controller electronics, battery contacts, battery wiring and solder joints, battery retention clips.

In addition, it would be totally unacceptable in a manned aircraft to have a situation where the flight controls could just stop working, they can easily do this in a drone simply because of interference, jamming, a receiver failure, a transmitter failure, a software crash, a transmitter battery failure, flying the **** thing behind a hill/tree/building and probably more events that I can't think of right now.

We're still in "basic quad" territory with all these possibly catastrophic failures. When we add more complexity with GPS and so forth, it gets worse not better. What happens when, just like my phone or car satnav, a reflection or other error makes the GPS report an incorrect position (and don't tell me your satnav has never shown you on a parallel road a few meters or tens of meters away)? Now our "smart" drone isn't so smart as it uncontrollably flies itself 20 meters in a random direction trying to get where it thinks it should be whilst we watch someone's face/car/grandma getting bigger in the smartphone app we're using as a flight control system. Even ignoring GPS failure modes, GPS is barely fit for aviation applications when it's working properly, there's a reason we're still not routinely flying GPS approaches in airliners and don't tell me anyone checks RAIM before flying their drone.

Then there's the pilot who has no training, no license and no medical.

Having two compasses, or two IMUs, when both of them are just being used to compute a single position and attitude solution is not redundancy. In fact, it can at times be worse than having just one since there's no third source to "vote out" the bad source, as evidenced by the "compass disagree" failures in the mavic that people have been having where it can revert to atti mode and throw out the GPS due to a noisy back compass despite still having perfectly good data from two sources (GPS and front compass).

Redundancy is two entirely separate, not interconnected systems that can do the same thing, they must not be interconnected or providing inputs to a single point of failure further up the chain and they must not be able to **** with one another no matter what fails or how bad things get - there's absolutely no redundancy whatsoever in a mavic, categorically none at all.

RTH is not a fail-safe, it's a fail dangerous. When the controller disconnects you have a pilotless aircraft doing something that's at best less dangerous than the alternative, but can only be described as safe in the loosest possible sense of the term. I am by the way a big fan of the feature - it is a great idea and works well if properly configured and in good conditions, but if you want to see how dangerous it can be you only have to look at all the crashes reported on here a significant proportion of which happened during an autonomous RTH.

If all of this sounds over the top, allow me to invite you to purchase tickets on my new budget airliner which will be coming to market soon and bringing ticket costs down to unprecedented value pricing by using a combination of off-the-shelf software from smartphones, mass produced Chinese hardware that works as it's supposed to at least 99% of the time and unlicensed pilots who learned by trial and error.

Finally, as I said, I'm wholeheartedly in favour of reasonably unregulated drone flying - it's fun and safe provided people don't get carried away thinking these things are idiot-proof or reliable. There's always a big enough idiot to break the idiot proofing (hint: it's normally you) and Murphy's law ALWAYS applies.

Nice post but checking RAIM would be a bit silly since the MP does have GLONASS ... the RAIM algorithms don't look at at that. And yes, the satnav on board is not separate receivers but one chip. If the chip dies you lose both GPS and GLONASS.

And while the MP may be subject to multipath, that risk goes way down as a function of its height above ground ... (not in the city of course when between buildings - but drones have no business there in any case ,,, or IMO).

Your thesis above about comparing to the very deep civil aviation requirements has a single major flaw:

Those requirements are for aircraft that carry people.

The requirements for, say, a privately owned Cessna flying VFR are far less than a commercial twin flying IFR which are far less than a corporate turboprop, which are less than an airliner in domestic airspace which are less than oceanic flying 3 or 4 engine aircraft which are less than oceanic 2 engine aircraft.

That's why your cheap airline analogy just doesn't hold water. Because ... it doesn't hold water.

Perspective wot. And compared to an unmanned aircraft that is somewhat restricted in mass (and therefore very limited in the damage it can cause on the ground), the requirements for a drone, recreational or commercial will never need to be as stringent as those for even a C-172 flown VFR. And that has nowhere near the computing, GPS/GLONASS, stabilization and so on that a Mavic Pro has.

You're also right about RTH. It should (simply put) be configured only for a max height return with OA off. And don't fly under things.

Comparing these realms has limits and the key thing to consider is not whether a drone has the backups of a 787 but what is the risk to people when things fail. And unless you're flying over a crowd (frowned upon) the risk is mostly to the owners pocketbook.

Up to the users to learn and teach and use above average judgement.

Here's a thought: for each one of us here there are 10 MP users out "there" who are flying under a lot of false assumptions ....

 

[BorisTheSpider]

I know of at least one likely battery detachment: Mavic Fly Away this evening. Shes Gone!

I kind of see where you're coming from about wing detachments. I'm not an aeronautical engineer but I do know that one "possible" (as in thinkable) failure like a spar bolt can't cause that, perhaps you have a point in that a single component that's unthinkable to fail like a spar itself is allowed to be a single point of failure, but like you said it's something where we cannot envision any reasonable scenario where it could spontaneously fail.

That can't be said of a quadcopter prop, prop mount, motor, ESC, CPU, battery, a solder joint or many other parts, although maybe you have a fair point about the battery contacts.

 

[AlanTheBeast]

Mavic, Phantom, Inspire are expensive, smart, incredible "pieces of technology", but definitely TOYS. And very dangerous toys.

A dangerous toy in the hands of a fool and a tool in the hand of someone responsible who practices safety.

The weak link is not the drone - it's the operator.

 

[AlanTheBeast]

I know of at least one likely battery detachment: Mavic Fly Away this evening. Shes Gone!

I kind of see where you're coming from about wing detachments. I'm not an aeronautical engineer but I do know that one "possible" (as in thinkable) failure like a spar bolt can't cause that, perhaps you have a point in that a single component that's unthinkable to fail like a spar itself is allowed to be a single point of failure, but like you said it's something where we cannot envision any reasonable scenario where it could spontaneously fail.

That can't be said of a quadcopter prop, prop mount, motor, ESC, CPU, battery, a solder joint or many other parts, although maybe you have a fair point about the battery contacts.

We don't _know_ that the battery detached. It's a top contending theory because almost everything else was ruled out. But there is no proof. And such a failure would be operator - not drone. If you latch that battery properly it is not coming out. If only one latch engages, then it is _plausible_ but not certain.

 

[BorisTheSpider]

That's a ridiculous statement. You know very well that there is not complete redundancy of all flight controls and flight surfaces on commercial and military aircraft.

That's quite untrue for civilian aircraft. Maybe there aren't always separate surfaces, but when there aren't there are always entirely separate hydraulic systems moving them. Quite often there are split surfaces as well, but at the very least there are separate actuators that can't jam each other and breakouts for a jammed control run.

For your definition of "redundancy" to be true, there would have to be two of EVERYTHING related to flight on every aircraft, including two separate and independent sets of horizontal and vertical stabilizers, and we all know that that isn't so.

That is exactly what we have on airliners - at least two entirely separate sets of flight controls with zero interdependency. Like I said, where it can be shown to be safe that may be achieved with a single physical surface which is movable by two entirely separate control systems but in that case you may as well have two surfaces.

As for the statement that having two compasses and two IMUs provides no redundancy because there is no third vote, that's an overly simplistic view. Certainly, greater redundancy is provided by having at least three independent systems because then a simple "majority vote" algorithm can be applied. "Majority vote" works in the case of either a "soft failure" of one of the units (i.e., fails in such a way that the unit is still operating but gives erroneous but possibly plausible outputs) or a "hard failure" of one of the units (i.e., fails completely and goes dead or outputs completely nonsensical data). Granted, if one has just two IMUs, it becomes difficult to impossible to detect a "soft failure" of one or more units, but a "hard failure" involving one of the units is easily detectable. There is redundancy. If you don't believe that, try telling me that you wouldn't prefer to have two GPS units on you rather than just one if you're wandering around alone and lost in the wilderness?

That's a fair point. I don't know how gracefully a compass or IMU hard failure would be handled by a mavic but assuming it was well handled you're quite right.

 

[BorisTheSpider]

Nice post but checking RAIM would be a bit silly since the MP does have GLONASS ... the RAIM algorithms don't look at at that. And yes, the satnav on board is not separate receivers but one chip. If the chip dies you lose both GPS and GLONASS.

And while the MP may be subject to multipath, that risk goes way down as a function of its height above ground ... (not in the city of course when between buildings - but drones have no business there in any case ,,, or IMO).

Your thesis above about comparing to the very deep civil aviation requirements has a single major flaw:

Those requirements are for aircraft that carry people.

The requirements for, say, a privately owned Cessna flying VFR are far less than a commercial twin flying IFR which are far less than a corporate turboprop, which are less than an airliner in domestic airspace which are less than oceanic flying 3 or 4 engine aircraft which are less than oceanic 2 engine aircraft.

That's why your cheap airline analogy just doesn't hold water. Because ... it doesn't hold water.

Perspective wot. And compared to an unmanned aircraft that is somewhat restricted in mass (and therefore very limited in the damage it can cause on the ground), the requirements for a drone, recreational or commercial will never need to be as stringent as those for even a C-172 flown VFR. And that has nowhere near the computing, GPS/GLONASS, stabilization and so on that a Mavic Pro has.

You're also right about RTH. It should (simply put) be configured only for a max height return with OA off. And don't fly under things.

Comparing these realms has limits and the key thing to consider is not whether a drone has the backups of a 787 but what is the risk to people when things fail. And unless you're flying over a crowd (frowned upon) the risk is mostly to the owners pocketbook.

Up to the users to learn and teach and use above average judgement.

Here's a thought: for each one of us here there are 10 MP users out "there" who are flying under a lot of false assumptions ....

Good post - you are correct, of course the risk is smaller than a big aircraft, but that's the point - because the aircraft, risk and price are small, so is the design safety. We started this topic talking about flying over towns and cities. I fully agree that flown away from groups of people, consumer drones are a small enough risk to be acceptable, in line with many other risks we all take every day. Although it's possible to hurt someone flying one well away from built up areas, it's very unlikely and is a reasonable, acceptable and proportionate risk. The problem is when people think their fancy piece of consumer electronics is suitable for flying over towns and I disagree with that.

 

[BorisTheSpider]

such a failure would be operator - not drone. If you latch that battery properly it is not coming out. If only one latch engages, then it is _plausible_ but not certain.

I like your posts and agree with them, at least in this thread so I hope we won't butt heads, I'm sure we largely agree.

But to pick hairs, systems that affect safety should be designed, as far as possible, to accommodate idiocy because that's what we're all capable of at times. If you can think the battery is latched, but it's not, that's not a safe system. Even then, those mounts are made of what? The same plastic as the props that keep this thing in the air. I'm sure it's very good, but it's not to be relied upon - it'll probably not fail on you, but it's not just you flying one of these - at some point, statistically speaking, it will fail - no problem when it's over an unpopulated area but no fun when it's over the middle of a large settlement.

 

[frankiez]

LOLOLOLOL

The Mavic is absolutely toy grade. Commercial and industrial multi-rotors start at $10k and up. There is nothing "pro" about the Mavic beyond DJI's marketing.

However, that's beside the point...

Flying over dense concentrations over people, property, vehicles, and radio interference is just asking for problems. I fly NEAR downtown and take pictures of downtown, but it do it from the safety of clear areas. A Mavic falling from over the height of downtown buildings onto a city street could easily kill someone.

As well as your Mavic suddenly becoming unresponsive to your commands due to unforeseen interference and actually flying off into that downtown that your filming. The possibility exists even though you thought you were safely far away from people.

Just as you can be the safest driver and suddenly you're killed due to a drunk driver.

 

[stratos]

The Canadian law said not over "populous" areas. What does that mean?

I took it to mean anywhere on the chart (VNC) that was in yellow. (Of course these charts are notoriously out of date where demographic symbols are concerned...).
View attachment 8944


But, with the new interim rules I'd say I can't fly laterally close to anyone, but I could fly 3 mm over their heads and homes without issue as long as they are not part of an "open-air assembly of persons"

That's how obscenely stupid the Canadian government is.

I saw the canadian new rules that they want to apply, seriously they are ridiculous and its even more ridiculous on how they plan to enforce them.
Like seriously, how will they stop people in their park having a pic nic brinding their mavics to take "dronies" I mean that is completely harmless!

 

[frankiez]

I too dread the day that an innocent bystander is critically wounded from a hobbyists drone. .

I read an article awhile back about a professional drone pilot who was filming a wedding and had put the drone down and apparently some kid got ahold of it and started flying it. He crashed it into people and seriously injured them. Needless to say, the drone owner was being sued.