Network Rail drone crashes into garden

[Sporty60]

Apologies if this has been posted elsewhere.

I expect that technology will reduce the chances of this type of incident occurring over time.

Should drones be subject to an annual service or certificate of airworthiness, I wonder?

I know nothing about them but do purchasers just take them out of the box and off they go.

Official report calls for drone restrictions after crash 'could have killed'

The drone was being flown for a Network Rail survey when it plummeted from 70metres into a garden

www.hulldailymail.co.uk

On the afternoon of December 2 last year the unmanned DJI Phantom 4 RTK was being operated in automated flight mode by Network Rail to survey track at Newtongrange in Dalkeith, Midlothian.

However one of the four propellers detached mid-flight and the drone rapidly descended from a height of 70 metres (230ft) where it fell into the back garden of a house nearby.

The Air Accidents Investigation Branch (AAIB) has investigated the incident and found a member of the public was 10 metres from where the aircraft landed....

 

[ExRes]

Presumably it must mean that type of drone is flying without ever having "demonstrated an acceptable level of safety", how?

 

[edwin_m]

On a quick read it appears there was a loophole in the old regulations, because the required 50m separation of drones and people could be in a vertical direction, with the risk that if the drone fails it could land on someone from a height of 50m or more. There have been no serious injuries from falling drones as far as I'm aware, but that may just be because they haven't been around long enough for a low-probability event to happen - there have been several near-misses. A transition to new regulations is now in progress, which if I read it correctly would forbid this combination of circumstances and require operators who can't be certain whether anyone is at risk to use a much lighter drone.

It's also mentioned that during its descent the drone was accelerating at a greater rate than free-fall, because the remaining propellers continued to turn when it was inverted. Since I imagine they have to include an orientation sensor already, perhaps they could shut off or even reverse the motors if it was detected to be upside down?

 

[JohnMcL7]

On a quick read it appears there was a loophole in the old regulations, because the required 50m separation of drones and people could be in a vertical direction, with the risk that if the drone fails it could land on someone from a height of 50m or more. There have been no serious injuries from falling drones as far as I'm aware, but that may just be because they haven't been around long enough for a low-probability event to happen - there have been several near-misses. A transition to new regulations is now in progress, which if I read it correctly would forbid this combination of circumstances and require operators who can't be certain whether anyone is at risk to use a much lighter drone.

It's also mentioned that during its descent the drone was accelerating at a greater rate than free-fall, because the remaining propellers continued to turn when it was inverted. Since I imagine they have to include an orientation sensor already, perhaps they could shut off or even reverse the motors if it was detected to be upside down?

There's different sets of rules depending on the type of drone and the type of use and CAA licensed drone operators are allowed to conduct flights over people so it isn't a loophole this drone was above someone and they'll be able to do so under the newer rules. The drone in the picture is the much bigger DJI Inspire whereas the drone that crashed was a Phantom 4, not as small as their Mavic range but not that big a drone either. As a non-professional pilot I wasn't allowed to fly over people under the old rules and under the newer rules I need to be further away as my drone is a legacy drone, from what I recall the vertical loophole only applied to properties where you could fly above them if the drone was 30m clear.

I'm not convinced about the drone accelerating when inverted as the drone would normally only be idling the props anyway and in videos of this type of drone crashing like this, they've been in normal free fall. I'm less familiar with the Phantom range but the Mavic range drones will power off the motors automatically when the drone is inverted and they'll also kill power if they detect a major problem so if they have an impact, they'll stop the motors.

Lighter drones aren't a fix either because they still present a risk not just from a physical impact from falling but there's other risk such as the props shattering at high speed which while not fatal can cause nasty cuts. What concerns me is that as the drones get better and cheaper they're becoming much more accessible and there's more risk of accidents. The DJI drones do come with some software restrictions built in so you can't fly the drone into an airport area without authorisation and there's various other no-fly areas but most of the regulations aren't enforced by the drone so if I wanted I could set my drone up in the back garden and proceed to break every rule of the drone code (fly above 120m, fly within 30m of property, fly over people, fly outwith VLOS).

I am a huge fan of my drone since it gives incredible views but I'm also very careful with it keeping it away from people and buildings which is safer for them and the drone but I worry they're going to need to enforce the rules more strictly within the drone software which will in turn impact legitimate users. Such needing a network unlock before flight which is an issue since flying in remote areas where it's safer often doesn't have a good data connection.

 

[edwin_m]

I see nobody's linked the report, so here goes: https://www.gov.uk/government/news/...dropped-to-the-ground-newtongrange-midlothian

The UAS, a DJI Phantom 4 RTK, was being operated in an automated flight mode to survey a railway track and surrounding infrastructure when one of the four propellers detached whilst in-flight. The aircraft rapidly descended from a height of 70 m (230 ft) where it struck the ground in the rear garden of a house. No persons were injured.

This investigation has reviewed the new UAS regulations introduced on 31 December 2020 concerning the safe overflight of people and data available to assist in risk assessments. Two Safety Recommendations are made to the UK CAA.

There's different sets of rules depending on the type of drone and the type of use and CAA licensed drone operators are allowed to conduct flights over people so it isn't a loophole this drone was above someone and they'll be able to do so under the newer rules. The drone in the picture is the much bigger DJI Inspire whereas the drone that crashed was a Phantom 4, not as small as their Mavic range but not that big a drone either. As a non-professional pilot I wasn't allowed to fly over people under the old rules and under the newer rules I need to be further away as my drone is a legacy drone, from what I recall the vertical loophole only applied to properties where you could fly above them if the drone was 30m clear.

I'm not convinced about the drone accelerating when inverted as the drone would normally only be idling the props anyway and in videos of this type of drone crashing like this, they've been in normal free fall. I'm less familiar with the Phantom range but the Mavic range drones will power off the motors automatically when the drone is inverted and they'll also kill power if they detect a major problem so if they have an impact, they'll stop the motors.

Lighter drones aren't a fix either because they still present a risk not just from a physical impact from falling but there's other risk such as the props shattering at high speed which while not fatal can cause nasty cuts. What concerns me is that as the drones get better and cheaper they're becoming much more accessible and there's more risk of accidents. The DJI drones do come with some software restrictions built in so you can't fly the drone into an airport area without authorisation and there's various other no-fly areas but most of the regulations aren't enforced by the drone so if I wanted I could set my drone up in the back garden and proceed to break every rule of the drone code (fly above 120m, fly within 30m of property, fly over people, fly outwith VLOS).

I am a huge fan of my drone since it gives incredible views but I'm also very careful with it keeping it away from people and buildings which is safer for them and the drone but I worry they're going to need to enforce the rules more strictly within the drone software which will in turn impact legitimate users. Such needing a network unlock before flight which is an issue since flying in remote areas where it's safer often doesn't have a good data connection.

Good to see some feedback from someone with actual experience!

I've read the relevant parts again and agree this operation would still be allowed under the new rules with an appropriate license. However the AAIB do have concerns about this:

Safety Recommendation 2021-024 It is recommended that, until an analysis of failure rates per flying hour has demonstrated an acceptable level of safety, the Civil Aviation Authority should consider prohibiting the overflight of uninvolved persons for those unmanned aircraft operating in the Specific category which rely solely upon their propulsion system for lift that would, following a failure of the propulsion system, impact the ground with a kinetic energy exceeding 80 Joules.

Also:

In January 2021, Network Rail precluded the use of DJI Phantom 4s in support of its survey activities. Furthermore, they advised that they intend to carry out trials using a UAS with a MTOM of less than 250 grams for when there is a need for UAS operations over uninvolved persons. The FMS is also being updated to provide a ‘risk map’ to include information on areas having known hazards, such as transmission masts that could affect UAS communications.

A reduction by a factor of about 5 in the mass of the drone in this scenario would certainly help, not only because the kinetic energy is proportional to the mass but also because a smaller and lighter object of similar shape is more affected by air resistance so should free-fall more slowly. Clearly the hazard isn't totally eliminated so some caution about flying above people would still be appropriate.

The comment about downward acceleration comes from the report, where they appear to have found evidence of that happening.

 

[JohnMcL7]

I think part of the problem is that even if the rules prohibit flying over people there's no way to realistically enforce that, I find it frustrating how often I see the current drone rules being broken.

I'll be surprised if they find the sub 250g any use because these drones are very intolerant of any wind as they simply don't have the motor performance needed to cope with them, many DJI Mini owners have lost their drones entirely because they were blown away by the wind. I have one if the slightly bigger Mavic 2 drones which has a lot more power and even it struggles in the wind at times when it's higher up, it can easily manage conditions that the smaller Mini/Spark series simply can't fly in. I think the 700g drone is more realistic but at that point the weight savings are much smaller.

I'm curious what their reasoning is for the drone accelerating downwards is but they don't give any explanation.

 

[DarloRich]

Advert shown at the foot of this page on my phone:

launch your drone career in 2021!!!!!!!!

 

[Geezertronic]

Advert shown at the foot of this page on my phone:

launch your drone career in 2021!!!!!!!!

Same here

 

[Ediswan]

I'm curious what their reasoning is for the drone accelerating downwards is but they don't give any explanation.

The AIIB report says the aircraft was "on it's back". If the motors started running...

I first met a four rotor drone doing an aerial survey long before they became well know. Later in the day it took off again, rose to about 10m, loud bang as one drive failed, flipped over, and drove itself into the ground with the other rotors still running.

 

[JohnMcL7]

The AIIB report says the aircraft was "on it's back". If the motors started running...

I first met a four rotor drone doing an aerial survey long before they became well know. Later in the day it took off again, rose to about 10m, loud bang as one drive failed, flipped over, and drove itself into the ground with the other rotors still running.

This one was missing a prop though so even if the motors were at full power (which they shouldn't have been), they couldn't have pushed straight down due to the unbalanced force and they should have just been idling - there would need to be a lot of power to make the drone accelerate noticeably faster than gravity. I think I've found the answer in some flight logs for similar crashes with DJI drones that showed impossible descent rates so I suspect it's the same error here, not a very common error unless the props are incorrectly installed or aftermarket ones are being used so not much out on there on faults like this.

Not that it in anyway diminishes the crash as the drones fall with plenty force on their own, this one at the Olympics comes to mind from a few years ago that thankfully just missed:

 

[SouthStand]

I think part of the problem is that even if the rules prohibit flying over people there's no way to realistically enforce that, I find it frustrating how often I see the current drone rules being broken.

I'll be surprised if they find the sub 250g any use because these drones are very intolerant of any wind as they simply don't have the motor performance needed to cope with them, many DJI Mini owners have lost their drones entirely because they were blown away by the wind. I have one if the slightly bigger Mavic 2 drones which has a lot more power and even it struggles in the wind at times when it's higher up, it can easily manage conditions that the smaller Mini/Spark series simply can't fly in. I think the 700g drone is more realistic but at that point the weight savings are much smaller.

I'm curious what their reasoning is for the drone accelerating downwards is but they don't give any explanation.

It's called gravity

 

[TPO]

I see nobody's linked the report, so here goes: https://www.gov.uk/government/news/...dropped-to-the-ground-newtongrange-midlothian



Good to see some feedback from someone with actual experience!

I've read the relevant parts again and agree this operation would still be allowed under the new rules with an appropriate license. However the AAIB do have concerns about this:

Also:

A reduction by a factor of about 5 in the mass of the drone in this scenario would certainly help, not only because the kinetic energy is proportional to the mass but also because a smaller and lighter object of similar shape is more affected by air resistance so should free-fall more slowly. Clearly the hazard isn't totally eliminated so some caution about flying above people would still be appropriate.

The comment about downward acceleration comes from the report, where they appear to have found evidence of that happening.

Another solution is to use a new generation of drone with much more redundancy built in. Typical consumer/starting commercial ("Enterprise") drones have a number of possible single-point failures (all of which are likely to result in the aircraft falling out of the sky) which include:
- loss/failure of prop
- loss/failure of an ESC (electronic speed controller)
- loss/failure of a motor
- loss/failure of the battery
- failure of the IMU (Inertial measurement Unit- basically the solid state gyroscope)

Some drones now have a level of duplication- for example IMU duplication or 2 batteries. But landing without one prop/motor/esc is more difficult on a quad..... there was a fashion for "emergency parachutes" but the failure/misbehaviour of the firing mechanism was implicated in a couple of other drone crashes so they fell out of favour somewhat.

The latest commercial quadrocopter drone from DJI (Matrice 300) can land in a controller manner on 3 props. (It also has duplicated IMUs and compass and dual batteries). Similarly, hexacopters can lose a prop (or motor/esc) and land safely, however they tend not to have much battery life. I've also seen a bespoke octocopter designed for mountain rescue (in UK) which could fly in wet weather and 45mph wind, but it had other downsides when compared to the DJI stable (also was same price-band as Matrice 300, very much a commercial drone).

Another enemy of drones is moisture (typically causes ESC or motor failure), again the latest generation of DJI commercial offerings (Matrice 300) is IP45 rated and will fly in wind of up to 33mph. But as ever, there's a cost associated with this and also Network Rail still to some extent are in the mindset of the past-but-one regs as they distinguish between 7kg and below and over 7kg and as a result I gather have not moved to the Matrice 300 (which has a MTOM of 9kg).

Mini2 is a good little drone and bags of stuff for the price-band but it won't fly in damp or wet weather and won't fly in much wind. (Although worth saying that the flyaways are more to do with inexperienced operators than the drone itself). There are some IP67 rated drones out there but they are a bit specialist. There's also the question about how long the CAA will leave the sub-250g hole unplugged.

My interest in this: I have a commercial PfCO and hold the GVC (General Visual Line of Site remote UAV pilot) qualification so I am ready to transition to an Operational Authorisation. I can assure you there's much more invovled once you make the move from hobbist to operating under CAA PfCO (Permission for Commercial Operations) or the replacement for the PfCo since 1 jan 2021 (Operational Authorisation).

TPO

 

[Bullfield WEst]

Advert shown at the foot of this page on my phone:

launch your drone career in 2021!!!!!!!!

Log in and it will go away... and probably "Return Home" when the battery dies. :-0

 

[DarloRich]

Log in and it will go away... and probably "Return Home" when the battery dies. :-0

Was logged in - advert today:

Become a drone pilot and transform your business!

 

[2HAP]

Another part of the problem is that drones are not required to be able to maintain flight with one engine/propeller inoperatve.

 

[WatcherZero]

Whereas helicopters are required to be capable of autorotating to land safely after a failure. Its essentially putting the gearbox in neutral and using the residual rotation to provide a slowly decreasing amount of lift while adjusting the pitch which allows a modicum of control over blade speed by changing the equilibrium point of the rotation/air resistance friction boundary.