Receive a GPS Signal onboard a Train?

[Lord of nothing]

Hi
The problem is very simple. I need to receive the GPS Signal and draw my own Map.
The Main Railway Track Operator refuse to give me the GPS Information of his Tracks.   
That mean I have to collect it my own. The problem on the new Cars who have Windows who have a Metal Coating on. My ordinary USB GPS Receiver dont work.

My old Receiver is an U-Blox who work with the Software perfect. I like there Software who work Stable up to 12h and more.

How about that: http://www.navilock.de/produkte/1299_u-blox-NEO-M8U/62756/merkmale.html?setLanguage=en
Is there a chance that the work better?
I think the Austrian Railway dont give me any Datasheet from the Windows Manufacture to see what kind of Metal the use.

[NivagSwerdna]

Active Antenna?

or try and extract the detail from OpenStreetMap?

[cdev]

Skytraq sell very sensitive and accurate receivers. In some ways, even more so than ublox.

Multi-constellation, multi frequency systems are coming to all GPSs soon which will work on a much wider range of frequencies and be inherently accurate to only centimeters without needing RTK. Even inside buildings or possibly even inside of some train tunnels. This is so people can be located in emergencies.

Did you receive any signal at all through the coating?

[voltsandjolts]

Buy a GPS logger with magnetic or sucker mount - stick it on the outside of the train car before you get on!

Edit: This was just a joke, hence the exclamation mark. This would actually be a dumb thing to do.

[Lord of nothing]

Did you receive any signal at all through the coating?

With my "old" GPS Receiver no i dont think so.
I will make a Test the next time when I board the Train it will be next week.

Skytraq sell very sensitive and accurate receivers.

Do the have a Software who I can "Control" there Receiver and Record the GPS Track?

Active Antenna?

Good idea but the most "Pc/ Laptop" Style GPS Receiver dont have an Antenna Port. The next problem how to power the Antenna? The "old" Trains have no Powersocket.

or try and extract the detail from OpenStreetMap?

The Data get often from Images not from a Real GPS Position. I would like use the information in the future for my Big Data project.
For example if the Train dont Stop at the Station A1, A2, A3 and Stop on A4 it have to be a Fast Train.
The same think for Statistic Data. If the Train is at Time X on Point the need at least Time Y to get to Station A5...
But something like this need Time to find a solution for.

Multi-constellation, multi frequency systems are coming to all GPSs soon which will work on a much wider range of frequencies and be inherently accurate to only centimeters without needing RTK. Even inside buildings or possibly even inside of some train tunnels. This is so people can be located in emergencies.

So for what kind of Freq are you talking about? Sat Based Nav. like GPS, Glonas, Baidu or Cell Phone, Wifi Based?

with magnetic

The Chassy is Aluminium. 

stick it on the outside of the train car before you get on!

Can you recommend some with

sucker mount

who stick on by 230kmh?

[NivagSwerdna]

who stick on by 230kmh?

That's an idea.  Stick yourself to the outside of the train 

I think active antennas are <20mA @ 5V.  Not a huge deal.

Depending on your software but is likely that you could use pretty much any module that has a SMA or equivalent powered connector to an external antenna.  i,e, the software probably can be configured for $GPGGA.   

In any case your accuracy might be limited by line of sight to the sky.

Which receiver do you currently have?

[cdev]

Please don't attempt to stick your antenna on the outside of the train car-

For one thing, magnets don't work with aluminum. Also, its likely against the rules.

Doing so is likely, if not certain to get you into trouble as well as potentially dangerous to you if something were to happen to your GPS.

Just don't do it. get a better antenna and accept that you need to improve the signal that comes through that window, or give up on recording the train route, and settle for the maps you get from books/online. (which I am sure are better than the ones you are likely to be able to record.)

I have almost never run into a situation with a modern GPS where it could not get a signal if I was near a window. Even a tinted window.

Metallic window films may attenuate signals but they never have completely blocked them for me.
Make sure the antenna is facing up and outward.

Skytraq's control center is similar to U-center on ublox, its used for configuring the GPS's adjustable parameters for different applications. All GPSs have settings - the technical name for them is "Kalman filter", which tell the GPS what kind of movement is most likely to be signal and not noise.

For train travel, certain conditions would apply - Train travel is likely to involve low speed horizontal movement with very little of the vertical component. You want to make sure the GPS Kalman filter is set to the right setting. Pedestrian or car settings are likely to work best. Avoid the settings for a stationary GPS, model airplanes and rocketry. Most importantly, use a good active antenna with a ground plane. The ground plane can be small but it should be at least twice the size of the antenna itself. If you have a 25 mm ceramic patch, use at least a 50 cm square ground plane for best results. If you are using one of those magnetic active antennas, attach it to some steel object thats flat and put that in the top of your backpack facing upward. Make absolutely sure the antenna is facing upward. It can't work properly if it is oriented incorrectly.

I hope this is helpful.

Again: Don't do anything if it requires you attempt to stick anything anywhere outside of your own luggage.

Don't attempt to put your data logger or antenna anywhere other than inside your own luggage unless you have the train operators explicit permission.

Your own luggage is your business but their cars and your behavior while on their cars is their responsibility and they are trying to protect you and everybody else..

That said, it also seems to me that the information you seek is almost certainly available in maps, etc.

Even if you feel it isn't, and you're right, that's likely for some reason. 

Unless there is some very large reason to not do so, put common sense above your natural desire to "know" everything and default to accepting their decision making process.

Seriously.. 

Also, its likely you just need a better GPS and /or GPS antenna to get the signals inside of the car through the tinted window, safely.

[Lord of nothing]

Stick yourself to the outside of the train

I am not a German.

Which receiver do you currently have?

http://www.navilock.de/produkt/60095/pdf.html?sprache=en
(I think there is no other with the "red" Mark.)

Please don't attempt to stick your antenna on the outside of the train car-

I am not Stupid. 

For one thing, magnets don't work with aluminum.

Right.

and settle for the maps you get from books/online

The Maps dont have a GPS Position in.

I have almost never run into a situation with a modern GPS where it could not get a signal if I was near a window. Even a tinted window.

I believe your never got on an ICE T! The are horrible. Very Bad Cellphone Signal. Outside 100% -> all Doors Close -> ~20% Signal (in the Railway Station  ).

Train travel is likely to involve low speed horizontal movement with very little of the vertical component.

for you 200km/h is low speed? Depend on the Track there is vertical moment.

In Italy to Train goes up to 300km/h:


How about an Pannel Antenna?

Don't do anything if it requires you attempt to stick anything anywhere outside of your own luggage.

You never saw me onboard a Train.  When I put of my VHF Antenna the look like that -> 

is almost certainly available in maps, etc.

I dont need a Map I need the GPS Position of the Track.

[cdev]

I'm sorry, when I said "low speed" I was thinking of trains that went much slower.

Also, some people use GPS in model rocketry, high altitude ballooning, etc. What I said still applies to the Kalman filter, train movement is unlikely to contain much of a vertical component compared to its horizontal component unless you are on a funicular train. This is to allow the GPS to reject "cycle slips" which might otherwise show up as sudden jumps in the track file you record.

Also, I've heard stories of people (on much older, slower trains, whose windows opened) who stuck heads, arms, etc. out of said open windows and were decapitated or seriously injured.

[Richard Crowley]

Apparently I am missing something here?
I can right-click on any spot on Google maps and it will tell me the Latitude and Longitude to six decimal places.
That would seem more than adequate for the kind of application proposed.  For example...

[Lord of nothing]

who stuck heads, arms, etc. out of said open windows and were decapitated or seriously injured.

sure but a small Antenna for example VHF/ UHF who is "mounted" on the Window isnt a Problem (during Summer).

I can right-click on any spot on Google maps and it will tell me the Latitude and Longitude to six decimal places.

Yes but I need a 3D Trail. Just the Data I receive are "open source".

[Richard Crowley]

...I need a 3D Trail.

Why?  Are you targeting offensive weapons?  How much does altitude vary?  A few meters?  What are the consequences of a few m up or down?[/quote]

Just the Data I receive are "open source".

If I understood what that meant, I would  probably ask "so what"?
Perhaps you have not explained the actual project sufficiently for us to understand what seem to be bizarre and unwarranted requirements?  You seem to be painting yourself into an impossible corner for no apparent reason.

[Lord of nothing]

Are you targeting offensive weapons?

...

How much does altitude vary?  A few meters?

Here in our Country we have something who Call Mountain who the Train goes up and down.

Perhaps you have not explained the actual project sufficiently for us to understand what seem to be bizarre and unwarranted requirements?

First I will draw my own Map because there are no official open source Data.
Later with the Data I will try to make an Big Data thing and collect the GPS Data of many People who want to contribute to see where the Train have to break and calculate the Time who is need to reach an Station.
With the hole stats I can make predictions.
If i find people near an Railway Track I can track trains by there Wifi Hotspot (who is cheaper than Optical Tracking).

[Richard Crowley]

Thank you for that explanation.  However it does not clarify the need for altitude data?  Do you anticipate data from people tunneling UNDER the tracks?  Or from people flying OVER the tracks? I would think that reports from any position within 10m(?) of a straight-line approximation of the rail path would be "valid".

The operators of rail companies here in this country are considering use of semi-automated drones to patrol and inspect every mile of rail and the right-of-way to either side of the rail line.  They received extra encouragement after a recent derailment spilled 100s of gallons of petroleum product along a major river.

[Lord of nothing]

However it does not clarify the need for altitude data?

The most Trains goes the Mountain up and down.  So I will show it in 3D where the Train goes.

I would think that reports from any position within 10m(?) of a straight-line approximation of the rail path would be "valid".

10m 
I know modern GNS Systems have an accuracy of ~1m.

The operators of rail companies here in this country are considering use of semi-automated drones to patrol and inspect every mile of rail and the right-of-way to either side of the rail line.

Until the goes to low and fly into the powerline... 
Here we have some special inspection Cars who get hand on a (Passanger) Train and the measure the Track and Powerline.

[Richard Crowley]

The most Trains goes the Mountain up and down.  So I will show it in 3D where the Train goes.

But altitude isn't that important, is it?  Are you anticipating the tracks moving up or down?  Surely if there were ANY vertical movement it would indicate a far more serious problem (earthquake, volcano, etc.).  Surely the only people who care about altitude for ground-level transportation are the people who design and build the roads (or rails or canals, etc.)  The people who travel on those ways don't care about altitude, do they?  Do they have any choice about altitude?  The only people who care about altitude are aircraft pilots (and the air traffic control network).

10m 
I know modern GNS Systems have an accuracy of ~1m.

Perhaps we still don't know what the purpose of your project is?  I thought you wanted some method of rail commuters reporting (and alerting others) to rail transport information.  Train location, arrival (or delay), incidents that affect schedules, etc.  The precise location of the rail line doesn't appear to be of any particular importance here?  You aren't expecting users to go out and survey the precise location of the tracks, are you?  That seems pretty unlikely.

Here we have some special inspection Cars who get hand on a (Passanger) Train and the measure the Track and Powerline.

But by the time that special inspection car has got to the location, the train has already passed through.  The purpose of the drones is to inspect the rails BEFORE any train traffic is assigned to go through that route.

[cdev]

Cartographic inaccuracy in the vertical plane is an important, under-recognized, issue along coastlines.

https://mobile.wunderground.com/hurricane/elevation.asp

Also, many newer cell phone GPSs are now RTK-capable and accurate to only a few centimeters or even better.

Of course, trains now have GPS trackers.. but many of them may not be RTK-aware and so only accurate to a few meters. Upgrades for train GPS trackers may be a low priority but they should be raised on the priority list because system managers need to know immediately if the track position shifts even a few millimeters downward.

they could offer immediate warning of any shifts in track position due to earth movements or heat expansion.

Of course, when cars allow the passage of RF statistical methods make it possible for transit systems to use aggregated data to generate precise, ongoing (almost realtime) measurements of any track movements (or gross elevation map inaccuracies) if any of the newly RTK-enabled riders phones receive signals at all. but they would then have to estimate the height of those riders phones..(in pockets a few inches above seat level) which introduces enough error to make the data much less useful.

They really should be able to be notified of any shifts that occur with millimeter accuracy.  that requires a sensor with a known height and position above the track.  it would actually be even better if two GPS receivers, one for each side of the track could be used.

[cdev]

"Moving base" RTK-GPS - is a technique which people have been using since around 2011 to determine the precise relation in space bwtween two points of known distance from one another..  I have fooled around with this in my backyard..

https://www.google.com/search?hl=en&as_q="Moving+base"+RTK

Short baseline RTK could give the exact location and position relative to one another of two GPSs on a drone's wingtips- combined with video - photogrammetry's "visual structure from motion" technology, or on track sensors which extended out behind the last car of a train to automatically survey tracks for minute movements such as would likely occur shortly before catastrophic track failures.

The most Trains goes the Mountain up and down.  So I will show it in 3D where the Train goes.

But altitude isn't that important, is it?  Are you anticipating the tracks moving up or down?  Surely if there were ANY vertical movement it would indicate a far more serious problem (earthquake, volcano, etc.).  Surely the only people who care about altitude for ground-level transportation are the people who design and build the roads (or rails or canals, etc.)  The people who travel on those ways don't care about altitude, do they?  Do they have any choice about altitude?  The only people who care about altitude are aircraft pilots (and the air traffic control network).

10m 
I know modern GNS Systems have an accuracy of ~1m.

Perhaps we still don't know what the purpose of your project is?  I thought you wanted some method of rail commuters reporting (and alerting others) to rail transport information.  Train location, arrival (or delay), incidents that affect schedules, etc.  The precise location of the rail line doesn't appear to be of any particular importance here?  You aren't expecting users to go out and survey the precise location of the tracks, are you?  That seems pretty unlikely.

Here we have some special inspection Cars who get hand on a (Passanger) Train and the measure the Track and Powerline.

But by the time that special inspection car has got to the location, the train has already passed through.  The purpose of the drones is to inspect the rails BEFORE any train traffic is assigned to go through that route.

[AG6QR]

Here's a photo of my GPS unit on board a high speed train, just east of Padova, Italia, headed toward Venezia.  Our speed was 132mph, or about 212 km/hr.  The other number shown are the heading of 071 degrees, vertical speed of 0 feet/second, and elevation of 24.9 ft.

I took this photo almost ten years ago, October 25, 2007.  The GPS is a Garmin GPS V, purchased in approximately 2002.  It is using the stock factory antenna, a quadrafilar helix.  As you can probably see, it is next to the window.

That GPS unit is antique by today's standards.  Virtually any modern GPS units gets better reception.  Other than placing the unit near the window, nothing special was done to make this work.  I was seated on the north side of the train.  The south side should have had a slightly better statistical chance of getting more satellites, but being on the "wrong" side didn't matter.

[Lord of nothing]

sry for the short replay.
The problem old trains dont use(ed) window with metal on.
Newer Trains +/- 2000 have mostly window with metal who block everything.
The newest should use Window "lte freq" optimized windows.

[Lord of nothing]

So now I am onboard the Train and there is quite no GPS Signal.
I record the GPS Signal from the Train ride if someone want analyse the Signal I can upload it to my Hoster when I am back home. (Depend on the Delay I get the connection or not...) 

[Lord of nothing]

So with hours without an GPS Signal now I switch the USB Port an what happed?
The bastard now received at least 2 Signals from GPS Sat.     

[Rerouter]

Most of the doors have a buffer rubber where they meet, i would have just picked the quiet door and poked through a pigtail antenna as they close if the carradge was really such a faraday cage.

Equally the same follows for the bending junctions, those areas have a rubber skirt and generally are not foil or metal wrapped, so sitting adjacent to one should improve any signals.

[Lord of nothing]

omg... I dont want get banned from Public Transport.
The Train are pressure sealed and the door is very solid.
There is no way to pock anything troth.

So 30min bevor arriving my first transfer station I got an 2D Signal.
5 GPS Signals at the moment.

Does anyone here have experience with the Ublox Chiptset and maybe know why the have some strange behaviour?
A Standalone GPS Device hat hat the benefit to get a fix and track the Sat bevor someone board the Train.

Will a helix Antenna bring any benefit or does the Ceramic do a better Job?
THX! 

[Rerouter]

did you have a full 3D fix at the station before boarding the train, most recievers can maintain a 3D fix, but stuggle to fix one in poor recption conditions, once 3d fix is made you can maintain 10m positioning off even 3 satellites, but in reality a modern gps / glonass / bidao reciever should have about 16 signals outside of city centers. allowing for very good positioning accuracy.

If you have an android phone, you can compare with something like the app "gps status" to see how well your reciever is working in comparison

Equally its ideal to have a receiever on for atleast 10 minutes before you put it in poor conditions, by then it should have enough of the alminac to manage.

Ceramic patch antennas are what i am mostly used to using.