Railway cables overpowered errant drone's compass and flung it back to terra firma

A commercial drone fell from the sky after a flight across a railway line threw its internal compass into confusion. The Aerialtronics Altura Zenith ATX8 craft crashed into bushes next to a railway line in October 2019, according to a recent Air Accidents Investigation Branch report. The craft's operator started its flight …

[Ian Johnston]

Order of magnitude error

Overhead wires for trains carry AC current at 25kV. Current draw runs into the thousands of amps when a train is passing by.

A really powerful UK electric locomotive is 5MW, which is only 200A at 25kV. The current draw never gets near a thousand amps, let alone "thousands". The conact wire is only about 120 square millimetres.

[John Doe 12]

Re: Order of magnitude error

Not ALL trains are running at 25kV AC. For example the Tyne and Wear Metro in the UK is 1500v DC and I remember as a trainee working in the substation watching the ammeters climb well past 1000amps.

I seem to remember that Holland is also 1500v DC which is why they bought locos originally built to work the Manchester-Sheffield route back in the early 70's.

[Nick Porter]

Re: Order of magnitude error

Southern 3rd rail is only 750V and a 10-car 444 will draw a whopping 5kA.

[Ian Johnston]

Re: Order of magnitude error

Of course, but the article specified a 25kV overhead system, and that's a couple of hundred amps tops, unless on HS1 which I checked and this wasn't.

[Ian Johnston]

Re: Order of magnitude error

Metro trains are just over 500kW per unit and therefore top out at just over 300A. The substation will have been feeding many of them through multiple feeders.

[Mips]

Re: Order of magnitude error

But EM interference is not news.

Back in ‘85 We had a project to install a mainframe in a site which backed onto the Liverpool/Southport rail line: 1500V DC. The EM survey showed huge induced voltage spikes as a train went past. This site was abandoned.

Somewhat later we had to abandon a project near Blackpool airport. This was blasted by the airfield approach radar.

There is nothing new under the sun.

[diodesign]

Re: Order of magnitude error

Thanks - that's been fixed. Please drop us a note via email [e-mail] if you spot anything wrong so it can be addressed immediately and not a day later when we get round to reading comments.

C.

[The First Dave]

"It is unclear whether three similar magnetometers would have been immune to the strong electromagnetic field generated by a passing train."

Sorry, but its perfectly obvious that ANY magnetometers would have been affected - it was the field being 'unexpected' that was the problem, not the readings, so three independent devices would NOT have helped.

[Anonymous Coward]

Unless they were sufficiently spaced that the localised magnetic field deviated enough between them to register a difference.

But I agree, it wouldn't have saved the craft; better software could have done.

[Anonymous Coward]

low voltage

You really don't want to have any electronics anywhere near active overhead railway lines.

In a previous job I remember we had some problems with electronics that was sited close to railway lines with overhead cables - it was the other side of the fence but kept getting weird glitches. Enquiries were made to BR engineering and they said that any induced voltages at the distances we were talking about "should only ever be low voltage, so there should be nothing to worry about".

It took a little while to realise that the "low voltage" they were referring to was according to the IEC definition, i.e. up to 1000V.

[My other car WAS an IAV Stryker]

Re: low voltage

In Minnesota, the light-rail Metro Green Line between downtowns Minneapolis and St. Paul runs on 750 VDC. It passes right through the heart of the main University of Minnesota campus and displaced normal vehicular traffic on what used to be quite the busy avenue.

From Wikipedia: "A particular area required exclusive attention and support, that being outside the University of Minnesota's biology department. Between Pleasant and Harvard streets there was 1/8" of tolerance between design and as-builds. Control factors used in the biology labs would be impacted by the electromagnetic radiation emitted from the nearby light rail. The 1/8" tolerance is said to cancel out any EMR that the light rail may produce."

The press during time of design and build said the tolerance was 1/4", and "cancel out" is a misnomer -- we all know EMI would follow an inverse-square law (radius from conductor center since it's a line, not a point source) and be reduced but never truly zero unless one uses a Faraday cage, but the building(s) in question were older affairs which couldn't be upgraded that extensively.

There have been no public reports of any issues during testing or full operation. Kudos to the contractors who pulled it off.

Of course, if they had gone with a different routing this wouldn't have been an issue(!), but that would have entailed a new bridge over the Mississippi River, a bridge/tunnel option -- U of MN opposed that also due to seismic-sensitive projects -- or sharing the brand-spanking-new I-35W bridge that was built to be ready for such a task (I may have mentioned the 2007 collapse before). Politics and engineering just don't mix.

[Roland6]

Re: low voltage

>You really don't want to have any electronics anywhere near active overhead railway lines.

Actually, you don't want electronics anywhere near an active electric railway line - overhead or third-rail.

Whilst the arcing between the power and train is one problem, another is the fluctuations in the earth the passing of a train causes.

From memory these were particularly bad on the Liverpool St. lines, where they were known to cause the signalling system to reset.

[Mark #255]

Re: low voltage

Actually, you don't want electronics anywhere near an active electric railway line - overhead or third-rail.

This is generally true.

It's also true that modern railway signalling systems are chock-full of electronics, and tend to be sited Really Quite Close (TM) to the rails.

So while you wouldn't want to, you find that if required, solutions are available.

[Roland6]

Re: low voltage

>"It's also true that modern railway signaling systems are chock-full of electronics, and tend to be sited Really Quite Close (TM) to the rails.

So while you wouldn't want to, you find that if required, solutions are available."

I would hope by now the solutions are pretty much either off-the-shelf or "you just do it that way", but back in the late 1970s when the first microcomputer-controlled systems were being put in place (on Liverpool St. we were pulling out Victorian mechanical and early electro-mechanical block signaling systems...), our understanding of such matters was significantly less - with even thyristor switches caused 'interesting' effects on sensitive electronics that happened to be on the same power infrastructure.

[werdsmith]

Re: low voltage

In my experience, my phone continues to work with data and voice when on the station platform with trains passing or seated on the train directly under these cables. I've not tried the compass there, next time I don't know when in the future, I will try it.

Aged about 10 or so we had a rail line passing the end of the school playing fields and on frosty morning we were guaranteed a show of fizzy blue arcs as the trains passed the spots where the cables are suspended.

[Roland6]

Re: low voltage

Talking of 'ancient' Signalling Systems, I see Network Rail has got around to replacing the 1970's London Bridge system with a "more reliable" modern system.

[ https://www.railengineer.co.uk/2020/05/20/south-east-london-lines-to-close-for-9-days-over-the-summer/ ]

Does any one know when (or if) the London Bridge Digico Micro16's were replaced.

[cantankerous swineherd]

Re: low voltage

fault with signalling system not necessarily local scallies nicking the copper then.

[Pascal Monett]

"over localised regions of the bridge below"

Okay, so you decided to launch your drone on a bridge, likely made of metal but not specified in the article, and next to an electrified train line.

Instead of launching from a field in the middle of nowhere.

Were you specifically testing the drone's ability to navigate in complicated electromagnetic environments ? Because even without a metal bridge, an electrified train line is going to have an impact on compasses, as you found out the hard way.

[johnfbw]

Re: "over localised regions of the bridge below"

Given the people concerned knew about the "Network Rail (Route Control and Air Operations)" (why does Network Rail have an air department - they can barely run the things on the ground!) and they were flying a £30k commercial drone I would assume they were either investigating the bridge or the rails so operation near them was essential

[Stuart Castle]

Re: "over localised regions of the bridge below"

https://www.networkrail.co.uk/running-the-railway/looking-after-the-railway/our-fleet-machines-and-vehicles/air-operations/

They have an air department because they use drones (and sometimes helicopters) to inspect the railway network without having to close parts of it so that engineers can wander up the track safely.

[John Brown (no body)]

Re: "over localised regions of the bridge below"

British Rail used to be co-owners (with SNCF) of the cross channel hovercraft service, which is arguably an "air"craft :-)

[Steve Aubrey]

Re: "over localised regions of the bridge below"

Oooooh, that's low.

[YetAnotherLocksmith]

Re: "over localised regions of the bridge below"

It is in law too: the captain and appropriate crew have pilot's licenses.

[Antron Argaiv]

Re: "over localised regions of the bridge below"

I would have assumed that a $30k "commercial drone", would have had some kind of EMC (Electro Magnetic Compatibility) testing before being sent out into the market. And that said testing would have included what's commonly referred to as "immunity testing" to validate its performance in the presence of electromagnetic interference. The emphasis here being on the magnetic part.

Well, back to the test chamber with it. Better luck next time. From the description of the incident, the compass wasn't the only thing affected. Apparently the control system doesn't like high magnetic fields either, thus that altitude loss.

Had I bought one of those, I would be contacting my lawyers.

[anonymous boring coward]

Re: "over localised regions of the bridge below"

"Had I bought one of those, I would be contacting my lawyers."

Yes, but for them $30k is "found in the back of the sofa" money.

[EvilDrSmith]

Re: "over localised regions of the bridge below"

The bridge may have been metal, but it only spanned two rail lines, so there is a good chance it was masonry.

Given the value of the drone and that they knew who to contact at Network rail, this sounds like the drone was being used for some form of survey probably on behalf of NR, or at least on an asset immediately adjacent to the rail line. The bridge presumably provided a (the best? the only?) vantage point for the drone pilot to keep the drone in visual sight at all times (which I think is a legal requirement).

[Annihilator]

Re: "over localised regions of the bridge below"

Presumably not any DJI type drone anyway - mine bitches about an unknown magnetic field if I've got so much as a keyring in my pocket when doing the initial calibration.

[YetAnotherLocksmith]

Re: "over localised regions of the bridge below"

If you put it on a big metal sheet, it'll just reorient the compass when calibrating, I expect, unless it is bright enough to realise the field strength is wrong.

[Annihilator]

Re: "over localised regions of the bridge below"

You have to rotate it several times through first the z then x axis before it completes calibration. Check this guy out - put your phone down away from the drone, take your watch off etc. It's unbelievably sensitive.

https://www.youtube.com/watch?v=ayPqOl7gvWo

[jake]

Re: "over localised regions of the bridge below"

You have to put your phone down‽‽‽‽

And Millennials play with these things? How does that work?

[sanwin]

"Electromagnetic interference has been blamed for a number of unforeseen problems in electronic devices over the years"

Who knew! Unforeseen only by designers who know nothing about EMI!

[Anonymous Coward]

I am being amazed these days at how people are apparently able to get advanced degrees in subjects related to electronic engineering without, seemingly, understanding how EM fields work. It's as if you need a specific degree in radio to understand stuff that in the Good Old Days* they used to teach at, at worst, A level.

*Insert 4 Yorkshiremen jokes, I can't be bothered.

[jake]

Indeed.

"It's as if you need a specific degree in radio to understand stuff that in the Good Old Days* they used to teach at, at worst, A level."

In the late 1960s a couple of 10 year old boys in Palo Alto decided that they would revisit the early days of radio for their Science Fair project. They built a couple of fairly high powered spark gap transmitters and matching receivers. Testing them between their homes (a couple blocks apart) on weekends and evenings was done in short sessions because their families bitched about radio & TV interference. They alternated who was transmitting, and cycled to a nearby park[0] to compare notes (local telephone calls cost money back then).

Come the day of the Fair, one was setup in their classroom, the other in the "Multi Purpose" room with the other displays. Everybody wanted to get a good look at their dangerous looking and sounding contraptions in operation. It took the police from Moffett Field NAS almost 45 minutes to show up and shut them down.

The resulting kerfuffle lasted a couple hours. Turned out that the Navy had been trying to find the source of the randomly occurring noise for a couple weeks, and were not amused. At all. It was a Cold War thing. However, no charges were filed ("kids will be kids!" was still an acceptable mantra). The boys were even allowed to keep their kit ... with the admonishment that they were not to ever turn it on again, in no uncertain terms.

I sometimes wonder what would happen if a couple kids tried this kind of thing today ... sadly, however, at that age they aren't taught enough of the basics to even contemplate the concept, much less attempt to implement it. Sad, that ... we've lost something as a society.

[0] Meadow Park, now called Ramos Park, if it matters to you ...

[John Brown (no body)]

Re: Indeed.

"I sometimes wonder what would happen if a couple kids tried this kind of thing today ... sadly, however, at that age they aren't taught enough of the basics to even contemplate the concept, much less attempt to implement it. Sad, that ... we've lost something as a society."

While I agree with you on the education (or lack thereof) bit, I do wonder just how much disruption that sort of fairly high powered spark gap transmitter might cause nowadays with so many people and devices relying on access to the airwaves. At the least, it might screw up everyone's home WiFi for a few blocks, maybe mobile phone signals, set off burglar alarms, screw up peoples satnavs, trigger WiFi enable heart pacemakers, OMG it's the apocalypse! Again!!

[cantankerous swineherd]

Re: Indeed.

oh dear, I appear to be in possession of a document relating to the commission, preparation or instigation of acts of terrorism.

[96percentchimp]

Re: Indeed.

"I sometimes wonder what would happen if a couple kids tried this kind of thing today ... sadly, however, at that age they aren't taught enough of the basics to even contemplate the concept, much less attempt to implement it. Sad, that ... we've lost something as a society."

You falsely posit a golden age when every 10-year-old was educated and motivated to produce innovative feats of engineering. I suggest that you're just a moaning old codger who looks at the past through rose-tinted specs. Then, as now, these bright kids represented a fraction of the total kid population, most of whom live unremarkable lives. Today, the bright kids have even more access to the knowledge they'd require to do this, but they're probably focusing their efforts on coding-based shenanigans.

[jake]

Re: Indeed.

"You falsely posit a golden age when every 10-year-old was educated and motivated to produce innovative feats of engineering."

Oh, horse shit. There was no golden age. However, there was a period of time where kids were actively encouraged to experiment with things that today are completely outlawed in the classroom.

"I suggest that you're just a moaning old codger"

Guilty!

"who looks at the past through rose-tinted specs."

Again, horse shit. I'm only a pseudo neo-luddite. The so-called "good old days" had plenty of bad to go without all of today's mod cons.

"Then, as now, these bright kids represented a fraction of the total kid population, most of whom live unremarkable lives. "

Of course.

"Today, the bright kids have even more access to the knowledge they'd require to do this,"

Access, yes. But actual hands-on school learning? Not so much.

"but they're probably focusing their efforts on coding-based shenanigans."

Drag and drop in a walled garden isn't coding.

[Anonymous Coward]

Re: Indeed.

Strangely, what the Register calls "Tat bazaar Ebay" is an immense facilitator for kids who want to do practical physics and electronics. Add to that Farnell Electronics/Element 14/ whatever.

Give an interested kid £100 and he or she can get a basic two channel scope attachment for a PC, a cheap PIC or equivalent development kit, assorted small motors, components and the like. Give them the price of an iPad and they can experiment with serious stuff.

Electronic components, electromechanicals and mechanical gear has never been cheaper in real terms, there has never been more information available in the form of data sheets, experiments and design information.

Look on YouTube and people are demonstrating how to recycle old engines, bicycles and so on into weird and sometimes wonderful creations that would inspire any kid with a bit of oomph.

What has unfortunately changed is that once upon a time those kids would be being kept quiet for hours every week learning and doing homework for a dead language maintained by the threat of the cane. Nowadays spending hours on social media is more entertaining and usually less painful.

So, my obvious suggestion? Governments should block Facebook and Instagram and bring in 10 hours a week of compulsory Latin. We'd soon have a generation of (self taught at school) scientists and engineers again.

[jake]

Re: Indeed.

Yeah, you're right. There were so many kids in the proto-SillyConValley in the 1960s sweating over Latin, Aramaic and Koine Greek and other medieval schooling practices (under constant threat of the cane, of course!), that it's a wonder that the computer revolution occurred at all.

Kids nowadays spending hours watching YouTube videos get so much more done. Why, today we have an App that will tell you if you have symptoms of Covid19! All you have to do is answer a couple of questions on your iFad, and there you go! Where would we ever be without such a marvelous invention? Ain't technology grand?

[Fursty Ferret]

I appreciate that you can't design out poor decision making, but this should have been taken into account in the drone's flight software.

It has three-axis gyro data available, and even a rudimentary architecture should have made the drone fail-over to this inertial positioning when the compass data became erratic and implausible. It doesn't need to know which direction it's going from a magnetic point of view, and I'm surprised that it even takes that into account for primary heading data. Basic stabilisation should, in my opinion, be solely from the IMU.

Magnetic heading can be used to compensate for gyro drift, but I'd have thought derived GPS track is likely more accurate anyway.

[Sam not the Viking]

Electronic compasses use gps and WMM (world magnetic model).

Both Android and iOS use the earth's magnetic field in their positioning systems.

https://www.ngdc.noaa.gov/geomag/WMM/uses.shtml

There are some clever people out there.

[GrumpenKraut]

Relocating office...

New monitors had to be bought anyway. Flat screens still a new thing and somewhat expensive. So management opted for (rather big) CRTs. New site near rail roads, so all CRTs where flickering like hell whenever a train passed, and that was all the time. And there is no practical way to stop the flickering (shielding with a thick iron box not being practical). CRTs were sold off for pennies, and flat screens bought.

Manglement ------>

[Annihilator]

Re: Relocating office...

I'm amazed you weren't just told to deal with it. That's the usual Manglement way.

[GrumpenKraut]

Re: Relocating office...

Work law trumps that. Fun detail: the cost-saving by buying CRTs where minimal, so effectively they paid twice.

[werdsmith]

Re: Relocating office...

A cathode ray being pulled around by scan coil magnets is a fun thing to play with, strong magnetic fields in the vicinity of the shadowmask would leave the shadowmask magnetised in places and distort the colour into some pretty paisley like patterns. There was a technique involved waving a strong degaussing wand in a circular pattern whilst withdrawing it away from the screen front to fix it.

I really miss the fun I had working in my youth and wish we still did that kind of work in this country.

[SImon Hobson]

Re: Relocating office...

all CRTs where flickering like hell whenever a train passed

Pa, that's nowt. Try welding current being passed through the frame of a steel portal building.

"A few" years ago, our offices were being expanded - the factory & warehouse kept moving along into new extensions, then office were extended into the newly vacated bits. The contractors decided on "belt and braces" so welded all the joints as well as bolting them. Cue 'kin big welding transformer with welding earth cable just clamped to the nearest bit of framework while the welder goes round all the joints. The magnetic field has "very interesting" effects on the CRT screen the other side of the wall - "wobbling" the picture right off the screen !

[John Doe 12]

Boeing inspired drone....

...with Max-8 style "technology" built in!!

[Oh Matron!]

Re: Boeing inspired drone....

Too soon!

[David Pearce]

Third rail

Third rail systems are DC and have traction return currents appearing everywhere that they shouldn't. They must have alarming effects on magnetic compasses nearby

[Outski]

Re: Third rail

Third rail in the UK is 750V DC, so (not that I can remember much shoolboy physics) wouldn't that be a much reduced effect compared to 25 kV AC?

[Anonymous Coward]

Re: Third rail

Magnetic field scales with current. Lower voltage means more current at the same power.

[Outski]

Re: Third rail

Thanks for that, have an early one ------>>

[ectel]

Re: Third rail

the Magnetic observatory at Heartland in North Devon was sited there specifically as it was as far away from railways as they could find

"The observatory was purpose-built for magnetic work, and continuous operations began in 1957, the International Geophysical Year (IGY). Hartland is the successor to Abinger and Greenwich observatories. The moves from Greenwich to Abinger and then to Hartland were made necessary as electrification of the railways progressed, making accurate geomagnetic measurements impossible in South-East England."

[Anonymous Coward]

Unpopular engineering pedant

...says that Cables are Insulated therefore have minimal field; whereas Overhead Line Conductors are only insulated by air :-)

[Richard 12]

Actual engineer

Insulation has almost zero* effect on the electric and magnetic fields.

It helps to do some research first. I recommend ElectroBOOM.

* At very short range it does affect the fields by preventing breakdown, arcing and of course crispy death. Magnetic fields can also be concentrated by use of ferromagnetic materials.

[vogon00]

Re: Actual engineer

And just to be different, I recommend avoiding ElectroBOOM, especially if you are trying to actually learn something about this 'Elec-trickery'* stuff. That said, his method of educating people can only improve the gene pool in the long run..

He's a good source of entertainment, but a shite source of knowledge.:-)

* Perfectly SFW, possibly even funny!

[Ian Johnston]

Re: Unpopular engineering pedant

Accurate engineering pedant says that the insulation makes no difference to the magnetic field.

[Anonymous Coward]

Re: Unpopular engineering pedant

You must be a mechanical engineer. One perhaps that has observed that the field from three phase cables is relatively small.

That's because from beyond very short distances, since the sum of the currents is zero, the magnetic field declines to a very small value. Very close to the cable your compass or magnetometer head is closer to one conductor than the others and the inverse square law kicks in.

It's the separation of the out and return currents (along with arcing) that causes much of the railway problem.

Some Swiss trains are three phase - two overhead and rail return - and this does reduce the far field a bit. You would think the small separation in third rail systems would reduce the far field, but the low voltage* means very high currents and this tends to counteract the effect.

*I do mean low voltage. As defined by the IEC.

[BenDwire]

Re: Unpopular engineering pedant

You've just reminded me of my first year at Uni, when one of my courses taught us to design electricity pylons such that the conductors didn't attract or repell each other too much, which would have damaged the insulators (glass or ceramic disks, if you've never looked up)

I went into software ...

[Anonymous Coward]

Re: Unpopular engineering pedant

Rightly downvoted due to rushed posting and not really making the point I was trying to make! The point I was trying to make is that Cable = insulated, usually buried or tunnelled. Bare conductors for overhead lines are known as just Conductors.

Buried cables, at least in the operation I'm part of, nearly all have a metal screen round the outside for earthing. The screen stops the electric field; obviously the magnetic one is still detectable at close range. 3 phase installations eliminate a good chunk of what this looks like at range.

Network rail feeds are typically two-phase installations plus a neutral for earth return, so can generate all kinds of weird and wonderful signals. There are other weird arrangements like a concentric 2-phase conductor out there. Not used so much today but a few older installs still have it.

[Anonymous South African Coward]

Encircle Heathrow (and others) with a 25kV line - will stop those pesky drones from flying!

[stiine]

The ones from Boeing or Airbus?

[cantankerous swineherd]

line of pylons seems to go to ground hereabouts...

https://goo.gl/maps/5fLnqybxCfsChvFg8

[hoola]

That is most likely a transition to underground cable so that it does not offend all the pesky building owners. It is interesting that in North Wales they are now burying some HV cables to remove pylons from the landscape.

The 3rd Rail trains are crazy as they convert the DC to AC and step up so they can then use a chopper type controller and the some gear as the 25KV overhead. That is way they make such a racket when they pull out of the station. When is was a kid (we had BR Southern Region in green pain) the trains all sounded like milk floats as the controller was nothing more than a series of giant resistors and some sort of switch to provide power control. You could often hear the bang as the driver took the power off and the contact dropped out. For all its critics, 3rd rail is very robust and presents far fewer problems than overhead, mainly because it is a dirty great rail rather than a flimsy bit of cable. It can ice up buy usually by that stage everything has stopped anyway.

I believe that there are speed limitations that preclude much more than 100mph running.

[Anonymous Coward]

I thought the problem with Southern was that the speed limitation is the amount of power you can get into a system running at only 700V or so. The DC-AC conversion is much more efficient than the old brute force motor controller gear, but even so the capacity is determined by a formula which has to take into account number of trains on the line, operating speed, stop and start time and thermal effects.

The dirty great rail has to be connected by cable to the power station, and that's another limiting factor.

I used to enjoy explaining to visitors why the power going into a plating bath was no more than that going into a washing machine, but one of them had 300 square mm cables bolted to substantial busbars and a rack cabinet full of transformer and rectifiers, and the other just had a 3 pin plug an an ordinary mains lead.

[VerySadGeek]

Ilegal ?

My understanding is that is not legal to fly a drone within a certain distance (50m ?) of the railway line.

[jake]

Re: Ilegal ?

It's also illegal for you to remove the copper wire & pipes from your local station. Doing so will (hopefully) get you a term in the local hoosegow However, the dude doing maintenance is allowed, nay, expected to do that very thing. He even gets paid to do it.

[Annihilator]

Re: Ilegal ?

That's true - and weirdly something that's not even mentioned in the Drone Code.

https://dronesafe.uk/wp-content/uploads/2019/11/Drone-Code_October2019.pdf