It likely will not come as a surprise to everyone, but I suspect that many people don't understand that the GPS System's primary customer is the Department of Defense. That means, among other things, that civilian users aren't guaranteed access to the system under all conditions.
I was curious as I read the article about PTC (November 2014) regarding the extent to which the railroads have taken into account the possible loss of GPS access. In other words, are there backup systems that would continue to provide the railroad safety functions even in the absence of GPS?
James W. Craft, Ph. D.
jwcraftca@gmail.com
Not quite sure under what 'conditions' you think railroads, as a strategic national asset, would no longer be given access to the GPS constellation.
It has been a very long time since 'military precision' has been encrypted -- and in any case it doesn't matter; the current NDGPS methods using assets like state GIS beacon networks and a large number of 'enabled' cell phones with appropriate chip support have made it cheap and, in many places, ubiquitous to have practical accuracy far better than that provided by the ICO constellation alone. (IIRC one issue in the past was that even at military precision it was difficult to resolve lateral location on multiple track... not a problem with modern methods, even the 'ad hoc' kind.) With the addition of some additional primary time standards in these networks, it is reasonably possible for modern ground-based 'networks' to provide adequate functionality for PTC without an actual satellite constellation -- requiring little more than the establishment of additional beacons 40-50 miles apart where existing ground-based coverage doesn't exist.
If you are concerned about some sort of hack or space-based attack taking down access to the GPS satellite constellation == even if the ground-based alternative was not fully in place at the time, there are other constellations (Galileo and GLONASS) that could be used.
Finally, even 'if all else fails' it would be possible to use an adaptation of the Canadian method which uses beacons on the locomotives and receivers that indicate mutual proximity of locomotives. That would at least preserve safety against collision between consists...
JAMES CRAFT It likely will not come as a surprise to everyone, but I suspect that many people don't understand that the GPS System's primary customer is the Department of Defense. That means, among other things, that civilian users aren't guaranteed access to the system under all conditions.
So the military would not let the railroads, who are moving all of their equipment, food, fuel and ammunition, plus supplying the national industrial base have access to GPS? Hmmmmmm
What they have now that has been working for the last 150 years.
Dave H. Painted side goes up. My website : wnbranch.com
I am confused by the notion the the mititary could deny access to GPS. The GPS sattelites just broadcast the time - continuously. All the smarts is in the receiver. It compares the difference in the time it receives from the various satellites and triangulates to figure out location. That's it. I suppose someone could shoot down the sattelites.
The backup is the fixed signal system plus dead reconning. If GPS went away long term, you could use track transponders with antenae on the locomotives.
-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/)
FWIW, the "which track am I on" problem is going to be solved by input from the train dispatching system. (Does not make me all warm and fuzzy...)
A key point will be the relative ease with which the GPS-based PTC can be dialed back or disabled (requiring a return to current methods of traffic control).
While I wouldn't want crews to be able to disable the system, I would hope that PTC would not be so deeply interlocked with the locomotive that the engine can't be moved unless PTC says so.
Larry Resident Microferroequinologist (at least at my house) Everyone goes home; Safety begins with you My Opinion. Standard Disclaimers Apply. No Expiration Date Come ride the rails with me! There's one thing about humility - the moment you think you've got it, you've lost it...
(1) The "P-Code" limiter has been turned-off for years .
(2) There are more "birds" in orbit now than ever before. There is less moving around of birds to suit the military at the exopense of the civilian user, Then there are the Russians (GLONAS), the Europeans (Galileo) and the Chinese (BeiDou) with their own constellations. The big deal challenge will always be relative precision. The human computer will still be called on to sort out the always present inconsistencies.
(3) It's more than GPS alone. Autonomous GPS is notoriously squirrely* and the HARN network plus the repeaters and VRS were never a sure fire stand-alone network. The various different systems use similar, but different applications to help build redundancy (or so we are told)...The amount of computer power to process gobs of recorded epochs in real time to get decent reliable results is huge. Ten years ago, surveyors couldn't get the precision they get now in real time and their equipment is far more involved than what PTC is at now on the PTC equipped trains. If it were GPS alone, every meet would be an alarm, especially where 13 foot or less track centers still exist.
(*) UP runs PMV cars over their lines plus CSX and others multiple times a year. Precision from one run to the next are not that great, but over time they have something they can live with for their purposes. Surveyors can tell you the difference between accuracy and precision - most others can't. That works against the PTC doomsayers and, oddly enough, proponents of black box technology. For all the money thrown at PTC, don't expect it to work to the expectations being touted. At least at first. This will be one VERY long process with lots of waivers until system reliability gets to a multiple sigma comfort level.
23 17 46 11
JAMES CRAFT It likely will not come as a surprise to everyone, but I suspect that many people don't understand that the GPS System's primary customer is the Department of Defense. That means, among other things, that civilian users aren't guaranteed access to the system under all conditions. I was curious as I read the article about PTC (November 2014) regarding the extent to which the railroads have taken into account the possible loss of GPS access. In other words, are there backup systems that would continue to provide the railroad safety functions even in the absence of GPS? James W. Craft, Ph. D. jwcraftca@gmail.com
when the airlines use GPS for their nav systems it's going to be a pretty sever national emergency to shut the system down.
You can lauch a Minuteman III missle form Minot AFB in North Dakota, send it ten thousand mils and it wil ldeliver up to 12 W88 warheads within 300 feet o fits target. AT 400 kilotons each close does not matter, but one would drop within 300 feet of the very center of center Kremlin. They use both inersiail and GPS guidancesystems. That is pretty darn accurate.
mudchickenFor all the money thrown at PTC, don't expect it to work to the expectations being touted. At least at first. This will be one VERY long process with lots of waivers until system reliability gets to a multiple sigma comfort level.
YES! Lots and lots of issues - like "how do I know the information I'm feeding the braking algorithm is any good?" PTC depends on lots of "commercial grade" data for safety critical work.
In the November 2014 trains magazine on page 29 is a box article stating the PTC systems may be subject to hacking. That i a worry some prospect. IMO the ACSES of Amtrak's NEC is a much better approach as the control uses additional track circuits overlaying the cab signal system. That is important due to planned speeds and multiple tracks with a fairly easy ability to add tracks.
Does the ACSES system allow for ATS as well ?
In my opinion, ACSES is goofy. They took the PRR 100 Hz cab signal and added a 250Hz cab signal overlay so they could get some higher speed signal aspects and still retain backward compatiblity with the original PRR system.
That part's not so bad.
But, it doesn't have predictive braking enforcement - particularly at absolute stop locations. Nor does it know about curves, bridges, etc, with speed restrictions, so, they added transponders screwed down to the track. But, wait! What about temporary speed restrictions and work zones. Let's add 220 MHz data radios (but don't use PTC data messaging protocol!)
So, now they have two sets of cab signals, transponders and readers, and data radio communication - and it's all integrated into the goofiest display ever. It shows governing cab signal speed and governing transponder and data radio speed, but the engineer has to look at both to see which one is lower. And, it doesn't show him at all where he is and what the upcoming restrictions are. He has to know and react accordingly.
Worst of all, the transponder stuff isn't in use except in a few select areas on the NEC (150 mph territory). The rest of the NEC is not PTC complilant.
For any Amtrak locomotives that might want to operate on the NEC and in I-ETMS territory - they'll need two sets of equipment. NS, CSX and Conrail are hoping Amtrak can use those data radios to sent I-ETMS messages to allow PTC operation on needed portions of the NEC w/o having install all sorts of stuff on a captive fleet.
Ed, your first sentence, "Most of the consumer GPS systems have a built in error factor...." gives me an explanation as to how the GPS system two different people were using to find my house before I moved told each one to turn west instead of east when they came to the street I then lived on. My address was 4778 West West Point Drive (it should have been named "Military Academy Drive"). They were driving on 4800 West Street (that IS is the name of the street); they had to call me and ask how to get to my house.
I also wonder if some map designers use an inaccurate GPS to show the location of house numbers. My house number is 584, and at least two maps put my house east of the street that divides the 500's and the 600's.
Johnny
Consumer GPS is only as good as the GIS map data the GPS vendors are supplied.
Never too old to have a happy childhood!
BaltACD Consumer GPS is only as good as the GIS map data the GPS vendors are supplied.
If it (GIS Data) screws up finding your house, then it's a slam dunk that it can't tell you a thing about the railroads in the area. Even the FRA's Tiger/ ArcVue data is a screwed-up mess (calling it a cartoon would be an improvement)
mudchicken BaltACD Consumer GPS is only as good as the GIS map data the GPS vendors are supplied. If it (GIS Data) screws up finding your house, then it's a slam dunk that it can't tell you a thing about the railroads in the area. Even the FRA's Tiger/ ArcVue data is a screwed-up mess (calling it a cartoon would be an improvement)
My consumer GPS doesn't show railroads on it's maps.
Deggesty Ed, your first sentence, "Most of the consumer GPS systems have a built in error factor...." gives me an explanation as to how the GPS system two different people were using to find my house before I moved told each one to turn west instead of east when they came to the street I then lived on. My address was 4778 West West Point Drive (it should have been named "Military Academy Drive"). They were driving on 4800 West Street (that IS is the name of the street); they had to call me and ask how to get to my house. I also wonder if some map designers use an inaccurate GPS to show the location of house numbers. My house number is 584, and at least two maps put my house east of the street that divides the 500's and the 600's.
As a pilot we almost always use GPS. However there are many restrictions. Most of my cohorts always back it up for several reasons.
1. There has to be at least 5 uninterrupted CPS satellites avaiilable the full trip time + two hours after. That has delayed a few trips. Aircraf that have inertial reference systems have much relaxed rules that are not known.
2. Selective availability is the defense department's ability to degrade accuracy to any error they desire. Supose a missle is coming toward the USA guess what will happen ?
3. During desert storm accuracy was actually improved because so many of the troops had only commercial units. Knew surveyors who worked fevershily during that time.
4. The unlikely problem of a coronal mass ejection ( CME ) aimed directly at earth being strong enough to permanently disable some or all satelites should be a concern.
5. Does anyone know if PTC systems will allow immediate default to present day traffic contol systems ?
oltmannd I am confused by the notion the the mititary could deny access to GPS. The GPS sattelites just broadcast the time - continuously.
I am confused by the notion the the mititary could deny access to GPS. The GPS sattelites just broadcast the time - continuously.
That is your answer right there. Change the time signal and accuracy is changed. The more you change the time the less accurate. Also GPS units do transmit position.
Every day at 0700 mountain time the GPS atomic clock and position is updated from Falon air force base ( almost forgot which one as it may have changed ) . Remember these clocks run at a different speed due to relativity factors. Some satellites are not in circular orbits and some polar orbits as well. complicating relativity. Time differences from different units establishes position. The math must have been wild ! It is a RHO - RHO - RHO slution much like loran.
Pictures & detailed drawings seen show a bank of corkscrew like antennas pointing at earth. Have no idea how directional they are. i
Last heard ( long time in past ) there are about 35 operational units and a couple older ones in a parking orbit as spares ?
One problem with GPS accuracy is atmospheric variances to the time it takes for the radio signals to get from the satellite to the receiver. These can be accounted for, to some degree, by putting a 2nd GPS receiver at a fixed/known location near to where the primary receiver is being used and have it broadcast on another frequency how much variance it is experiencing so the primary receiver can subtract that variance from its calculated location to give a much more accurate indication of location.
But without a very accurate and constant internal clock oscillator in the GPS receivers the calculations will be in error most of the time. Most "consumer" devices do not have a super accurate and temperature compensated oscillator in them ($$$$$) and so if you put a consumer grade device someplace and do not move it for a time, you will find that it will claim that it has moved continuously during that time. I have used 3 different GPS "consumer" platforms (DeLorme, Garmin and my Samsung Galaxy Nexus smartaleckphone) and none of them could guide a honeybee to a flower and back. I can sit on my porch and plot any of them showing that I am in the middle of the street, then in my neighbor’s living room, back across my porch and into the backyard, behind my other neighbor’s garage and then back into the street, down the street and then reverse and go behind the house across the street and back to my porch and then to my neighbor’s kitchen, and… well… I think you get the idea. It does “AVERAGE” to showing that I am near my porch, but it takes HOURS to get enough readings to average out to “close” to where I really was sitting.
If a commercial grade GPS receiver were in a locomotive, and you knew where the RR had tracks, then you could conclude that the locomotive was on tracks within half a city block of its actual location, but you could not tell which track it was on if there were more that one track in the general ROW. You might think it is on track #1, or possibly on the access road next to it (or the cornfield opposite that), and then it might show it travel across track #1 to track #2 (oblivious to the fact that there are no switches in the tracks to do so!) and into the creek on the other side, but don’t worry, it’ll soon get back at least near one of the tracks if you wait a short while (none the worse for getting its wheels wet in the creek).
The RR could install the secondary GPS receivers every few miles to transmit the atmospheric induced error info so the locomotive receiver could get better indications of where the locomotive is, but that seems to be an overkill in areas where there is only one track and only marginally effective to tell which track a train is on in multi-track territory.
Semper Vaporo
Pkgs.
Semper Vaporo The RR could install the secondary GPS receivers every few miles to transmit the atmospheric induced error info so the locomotive receiver could get better indications of where the locomotive is, but that seems to be an overkill in areas where there is only one track and only marginally effective to tell which track a train is on in multi-track territory. Forgot about that. You are actually on the right track. The FAA has installed broadcast units at some airports that give a very precise signal of their location. it is called differential GPS. Enables zero - Zero approaches without an ILS transmitter. Talk about pucker factors ! !
Forgot about that. You are actually on the right track. The FAA has installed broadcast units at some airports that give a very precise signal of their location. it is called differential GPS. Enables zero - Zero approaches without an ILS transmitter. Talk about pucker factors ! !
blue streak 1 oltmannd I am confused by the notion the the mititary could deny access to GPS. The GPS sattelites just broadcast the time - continuously. That is your answer right there. Change the time signal and accuracy is changed. The more you change the time the less accurate. Also GPS units do transmit position. Every day at 0700 mountain time the GPS atomic clock and position is updated from Falon air force base ( almost forgot which one as it may have changed ) . Remember these clocks run at a different speed due to relativity factors. Some satellites are not in circular orbits and some polar orbits as well. complicating relativity. Time differences from different units establishes position. The math must have been wild ! It is a RHO - RHO - RHO slution much like loran. Pictures & detailed drawings seen show a bank of corkscrew like antennas pointing at earth. Have no idea how directional they are. i Last heard ( long time in past ) there are about 35 operational units and a couple older ones in a parking orbit as spares ?
Points fogotten. Also had edited original post
The DOD signal uses an architecture ( Is changed periodialy ) that is completely different than civilian signal that is not changed so civilian signal can be degraded without changing DOD signal.
FAA was aware of this so they have installed some low power limited range transmitters at certain critical airports so selective availability will not shut down air traffic. Caused quite a stir at first at DOD.
Modern GPS receivers have Fractal antennas that can receive almost unlimited number of signals. ( Why current receivers mostly do not have a separate antenna ) Just depends on computer power in receivers. That is why foreign signals can be processed as well and are used. But that gives the possibility that those signals could be spoffed so causing mischief. As far as known Russia, China, Japan, France now have some form of GPS. Others ?
Latest generation receivers do not need an internal oscillator as the GPS transmitters all broadcast a signal. The first transmitter received usually becomes the master oscillator until it drops out of range. s
All this is public knowledge but this poster's reluctance comes from what is not known and whatever is not public knowledge ? ? !n
A few comments:
The "corkscrew" antennas are nire formally known as "helical" antennas and were invented by John Kraus after he attended a lecture on the just invented travelling wave tube. The antenas are inherently circularly polarized and doe provide directionality. They are also inherently broadband, which helps in minimizing distortion of the modulation GPS uses for trasmitting timing information (phase shift keying using direct sequence spread spectrum).
The "ideal" receiver for something as large as a train would be a phased array system capable of forming many "beams" that would be steered to where the satellites are expexted to be. This would provide better signal to noise ratio, some immunity to spoofing and some immunity to multipath. The downside is that the antenna would need a more complicated phasing system to preserve the timing info (this could be done with digitl signal processing).
Ionospheric disturbances will affect propagation velocity (which is way shortwave radio works for long distance communication) of the GPS signals, which messes up timing, which then messes up position accuracy.
Finally, most land based ICBM's use purely inetial guidance, to get the 300 foot accuracy requires a detailed gravity map for quite a distance from the launch site. SLBM's use astro inertial navigation, using stars to fine tune the inertial guidance. This is why some astronomers have been funded by the USN for improving precision of star position measurements. The increased accuracy can allow for warheads to be targeted at a crater left by a detonation a few minutes earlier, which allows for assured destruction of a deeply buried target.
- Erik
ERIK: Thanks for the antenna information. Had not heard some of that.
GPS and its functions as noted by Dr, Craft can be particularly problematic. Sunspots and their atmospheric disturbances cvan be very troublesome to electronic circuits. Their strength varies in their effect on this planet...
In September of 1959 this was a Sunspot whose radiation was so strong it effected the American Telegraphic Net, at the time. That was a system that was pretty simplistic in its functional hardware... These days the Radiation from a Sunspot can be strong enough to damage circuits and cause all kinds of electronic mischief...Computers are operational on many kinds of vehicles, and their shut down can cause all sorts of problems...To the extent that many systems might need to be 'shielded' (as in a Ferriday box) to maintain their integrity for operation.
Here is a linked website that might explain some of the functions and operations to some of us who do not have the technical expertise in GPS. @ http://www.gps.gov/
samfp1943 In September of 1959 this was a Sunspot whose radiation was so strong it effected the American Telegraphic Net, at the time. That was a system that was pretty simplistic in its functional hardware... These days the Radiation from a Sunspot can be strong enough to damage circuits and cause all kinds of electronic mischief...Computers are operational on many kinds of vehicles, and their shut down can cause all sorts of problems...To the extent that many systems might need to be 'shielded' (as in a Ferriday box) to maintain their integrity for operation.
I presume you were referring to the "Carrington Event" of 1859.
Computers on vehicles will likely not b affected by such an event, as the electromagnetic interference is a very low frequency needing a several mile long span of wire to generate the damaging voltages. The danger is worst where the crustal conductivity is low, so would be more of a problem in the Great Lakes area than the Gulf Coast.
The ionic dsturbances of a Carrington Event would mess up GPS reception and would likely cause damage to the satellites (though I would assume that the DoD wouldmake sure that such a strategic asset was properly radiation hardened).
erikemThe "corkscrew" antennas are nire formally known as "helical" antennas and were invented by John Kraus after he attended a lecture on the just invented travelling wave tube. The antenas are inherently circularly polarized and doe provide directionality. They are also inherently broadband, which helps in minimizing distortion of the modulation GPS uses for trasmitting timing information (phase shift keying using direct sequence spread spectrum).
Here is a patent for a dynamically reconfigurable helix that might have some benefit for railroad applications...
erikemThe "ideal" receiver for something as large as a train would be a phased array system capable of forming many "beams" that would be steered to where the satellites are expexted to be. This would provide better signal to noise ratio, some immunity to spoofing and some immunity to multipath. The downside is that the antenna would need a more complicated phasing system to preserve the timing info (this could be done with digitl signal processing).
An associated reference is this .pdf from the GNSS community.
Additional points that may need elaboration and / or correction from some of our surveyors such as MC.
11. The differential GPS transmitters located at some airports transmit both location and elevation so aircraft and other receivers can adjust.
12. This leads to the differences in elevation and altitude. Ground facilities are on elevation but air is based on indicated altitude.
13. Often they are synomous but there canl be differnces. Altitude is based on mean sea lever with the standard temperature of 59 F ( 15 degrees C ) and barometric pressure of 29.92 inches. For each 1000 ft of elevation the temperature decreases 2 degrees C. That introduces an error in altitude indications with the further from standard temperature the more altitude diifference. Canada requires altitude corrections higher for minimum enroute altitudes when low temperatures are below about -10C.
14. In the 1990s GPS observations in the US noted errors in position for some airports that were traced to errors in published elevation. Position is derived from elevation inputs from GPS as well aa the regular position signals which determine elevation. Some kind of computational feed back loop is used. That is one reason it takes a gps receiver longer to boot up when last known position more than 60 N miles from last position..
What was found after careful re surveying that 3 to 5 feet of elevation errors were found of touchdown zones. Aircraft antennas are calibrated from location to landing gear vertical distance. Scary were a few Mexican airports that had hundreds of feet elevation errors. Data bases for air and ground facility charted changes are revised every 28 days on a Tuesday. That made for hours of inserting revisions some Tuesdays as " do not use until ___ " messages were noted .
15. Complicating all this are
a. the earth is not a sphere not even a flattened shpere. Obblake sphereoid ( sp? hate no spell checker ) .
b. Gravity effects elevation especially at seaports that may have a quick fall off or a mile long shelg.
c. So what is mean seal level ?
16. So airplane flying in cold northern climes may have an altitude of 29,000 and elevation of 26,000 feet. Have observed in Brazil altitude of 29,000 ft and elevation of ~ 34,000 feet.
17. This is why the RRS are doing multiple surveys of their tracks to get actual elevation of the antennas which probably are mounted at same height above rail that loco antennas will be located ?
Probably will think of more item later.
The old saw " a little knowledge is a dangerous thing " applys. If you know that you only know a little that can make one very wary of the unknown. Hence my reluctance of seeing GPS used for PTC.
Our community is FREE to join. To participate you must either login or register for an account.