BaltACD Semper Vaporo rdamon I thought this was an interesting commercial when I saw it .. http://www.ispot.tv/ad/7mRw/ibm-analytics-railroad Hmmm... 230,000 cars total. Even if we assume that 130,000 are parked someplace all day, that leaves 100,000 in transit every day. Each car has 8 wheels (at least around here i have never seen any 4 or 12 wheel cars), so there are 800,000 wheels to be tested each day at least once. But they said in the video that they analyse 100,000 datapoints each day. They are only testing 1/8th of the possibly defective wheels each day. I suppose that is better than not analysing any wheels each day, but it seems a tad on the low side. Maybe it is early in the development stage and more detectors will be deployed in the future and better coverage will result. Note - IE11 plays the IBM commercial - Firefox gives the skin error.
Semper Vaporo rdamon I thought this was an interesting commercial when I saw it .. http://www.ispot.tv/ad/7mRw/ibm-analytics-railroad
rdamon I thought this was an interesting commercial when I saw it .. http://www.ispot.tv/ad/7mRw/ibm-analytics-railroad
I thought this was an interesting commercial when I saw it ..
http://www.ispot.tv/ad/7mRw/ibm-analytics-railroad
Hmmm... 230,000 cars total. Even if we assume that 130,000 are parked someplace all day, that leaves 100,000 in transit every day. Each car has 8 wheels (at least around here i have never seen any 4 or 12 wheel cars), so there are 800,000 wheels to be tested each day at least once. But they said in the video that they analyse 100,000 datapoints each day. They are only testing 1/8th of the possibly defective wheels each day. I suppose that is better than not analysing any wheels each day, but it seems a tad on the low side. Maybe it is early in the development stage and more detectors will be deployed in the future and better coverage will result.
Note - IE11 plays the IBM commercial - Firefox gives the skin error.
Ah! I had not noted the single carrier distinction... that does make a big difference, even if it were one of the largest ones that would inprove rate considerably.
Semper Vaporo
Pkgs.
Semper Vaporo rdamon I thought this was an interesting commercial when I saw it .. http://www.ispot.tv/ad/7mRw/ibm-analytics-railroad Hmmm... 230,000 cars total. Even if we assume that 130,000 are parked someplace all day, that leaves 100,000 in transit every day. Each car has 8 wheels (at least around here i have never seen any 4 or 12 wheel cars), so there are 800,000 wheels to be tested each day at least once. But they said in the video that they analyse 100,000 datapoints each day. They are only testing 1/8th of the possibly defective wheels each day. I suppose that is better than not analysing any wheels each day, but it seems a tad on the low side. Maybe it is early in the development stage and more detectors will be deployed in the future and better coverage will result.
The ad only mentions a 'single' carrier, measuring 100K data points - not all carriers. So to extrapolate that to the entirety if the rail network is in error.
Never too old to have a happy childhood!
dehusman You don't need to test every car every day, just like you don't take your car into the shop every day to get the tread measured and the tires inspected. Cracks take a while to propagate. You just need to test them enough that you can catch them between when the crack is big enough to detect, but small enought that it won't cause a catastrophic failure of the wheel. Most testing is done on an interval with a frequency less than the time it takes to fail (e.g. journals and rails).
You don't need to test every car every day, just like you don't take your car into the shop every day to get the tread measured and the tires inspected. Cracks take a while to propagate. You just need to test them enough that you can catch them between when the crack is big enough to detect, but small enought that it won't cause a catastrophic failure of the wheel. Most testing is done on an interval with a frequency less than the time it takes to fail (e.g. journals and rails).
Is current testing sufficient to prevent all wheel and rail failures, or is more testing needed to accomplish that?
Hope IBM's testing is better than displaying the video!
Error loading skin: http://p.jwpcdn.com/6/12/skins/bekle.xml
Dave H. Painted side goes up. My website : wnbranch.com
The UP has the "crack house" at N Platte where every inbound wheel on the westbound side is scanned as it rolls through a scanner building, looking for cracked wheels.
In addition there are the hotbox and dragging equipment detectors, wide load detectors, wheel impact detectors, high water detectors, bridge fire detectors, etc. Throw in event recorders, TIR, taped communications, signal logs, etc. and there is a whole bunch of data going on out there (and has been for a while).
csmith9474Please, no more cameras. Plus part of that responsibility would likely fall under my craft. We have installed some experimental sites that have an array of cameras and sensors that give a real time view of trains as the enter or leave a yard. They detect defects in wheels and couplers, among other things.
Outsiders have no understanding of how fully the multitude of defect detectors inspect the cars in trains as they pass - the hot box detector that has been around in one form or another since the 1960's is just tip of a veritable iceberg of of conditions that detectors have been designed to ferret out and report to responsible parties - both on trains and in inspection centers, with comparative readings being made between detectors on the train's route to highlight trending conditions.
rdamon I agree .. purpose built scanners are much better. I was just thinking how visual images could be used for discussion in this thread. I am sure those color bars on the sides of the cars in the 70's and 80's could be a real pain.
I agree .. purpose built scanners are much better. I was just thinking how visual images could be used for discussion in this thread.
I am sure those color bars on the sides of the cars in the 70's and 80's could be a real pain.
On gons the 70's era ACI bar codes lasted until the first load of hot steel slabs were loaded in the cars and burned the bar codes, as well as most of the paint on the cars off.
Cameras are a really bad way to scan, there are so many better ways and things to scan. To put it in hisotrical perspective, they hsed to have cameras beside the tracks scanning trains. Then camee AEI readers and they were 10000 times better at scanning cars at speed any time of day in any weather. We used to have humans scan trains for hot boxes and dragging equipment. Then the railroads put in thermal scanners and dragging equipment detectors that function way more reliably than somebody on the rear car of a 100 car freight who might be watching 50% of the time.
Actually railroads do have camera's on the front and rear of some trains, the TIR's. They just don't have methods of accessing the video real time. Once again, a solution that has a lot of the infrastructure in place (the lack of being the downfall of many of the ideas that are thrown against the wall on this forum). Engines already have power, transmitters and a physical location established. The real challenge becomes how to safely give access to the crews.
rdamon I could see an application where there are fixed cameras along the RoW. A central dispatcher could use this to do a remote roll by inspection. A series of camera locations could be used to allow a comparison to be done to detect irregularities. A baseline image could be set at the origin and using some sort of automated image comparison application, potential differences could be flagged for review at the remote center. In addition to visual images additional data like high and wide, dragging, smoke detection and other metrics could be tagged with video. This may help if there is a defect 8,000 or more feet back. And then the railroads could get suckers like us to pay for a live feed into our homes.
The various defect detectors that are required by the FRA to be at a maximum 50 miles apart, but, at least on my carrier are about ever 15 miles apart do a much more accurate job of 'scanning' the trains for specific defects - defects that would not necessarily be visible on a video feed. Additionally, passing and/or overtaken trains as well as MofW personnel also perform visual inspections of each others trains and communicate the results over the radio.
jeffhergertI've often thought a wide angle camera on the eot would help for back up moves.
I'd be happy if they stuck one on the long hood end of an engine. I know, some shortlines/industires do just that. Be nice if we had that luxury.
It's been fun. But it isn't much fun anymore. Signing off for now.
The opinions expressed here represent my own and not those of my employer, any other railroad, company, or person.t fun any
I've often thought a wide angle camera on the eot would help for back up moves. Right now there are limited times when a train can back up on a main or controlled (signalled) track without a person protecting the move. Specific requirements such as no crossing, public or private, will be fouled, etc. must be met. Even then, on a fenced right of way (fencing in the sense of a farmer's fence, not a chain link security type fence) doesn't exclude the possibility of a tresspasser or some other one in a million type obstruction from fouling the back up move.
To expand what Dave said, you would be surprised how much of a train you can't see from one spot. Even on a curve. If you really want the train to be scanned by a camera I think you would be better off equipping wayside detector locations. Still, to be of use in real time each train would require someone to actually watch each train. Recording a defect or trying to detect "passengers" doesn't do much when someone finally reviews it a day later.
Maybe computer technology could be developed, unless already exists, to "view" the images for defined events. (Events isn't the word I wanted to use, but I can't think of a better one right now.) It could then signal for a human to review the images and take any needed action.
Anyway you do it, it probably will cost a lot more than it benefits.
Jeff
Lots of logistical problems for no real return. Power source, secure attachment, interference with safety devices, bandwidth to transmit, how does it work at night, in the rain, in the snow, how to keep the lenses clean, how to keep them from being stolen, who puts them on, who takes them off, how do you associate a camera with a train with a location in the train.
Even if you solve all that, what have you got? Video. That doesn't detect or interact with anything. In order to be of use somebody has to watch it constantly. If they don't then you don't catch anything. Any camera close enough to the side of the car not to be a clearance hazard will have an extremely limited range of vision.
If something happens to the 3 cars ahead of the camera, whoever's watching will be all over it. 20 cars ahead of the camera, not so much.
If y'all just gotta stick a camera someplace, plut one on the EOT. It already has a power source and a transmitter, and accidently might just have a useful purpose.
There'd be an awful lot of lenses that need replacement after being spray painted black.
And as good as a GoPro camera is at taking hard-knocks, I doubt if they'd withstand RR knocks. The company I worked for got into the RR telemetry business many years ago and sent the Metrology Team to a RR to measure the G-forces associated with railroading. They took the same measurement instrumentation with them that was used for the Apollo and Space Shuttle rocket measurements... had to come back home and get equipment that could measure higher G-forces.
Cameras don't necessarily mean someone has to be always watchng them. They could be like store cameras and might even deter some of the foolishnesss and vandalism that now goes on.
CSSHEGEWISCHThe cameras may see a mid-train derailment as it happens, but what could be done about it?
And if they can make workable small cameras for space probes and landers, they can make cameras that work in "railroad" environments. Of course, the railroads may not want to pay the prices for those cameras..
Like libertarianism and Marxism, this is an idea that sounds better on paper than in the real world. I am assuming that monitoring these multiple cameras would be an additional responsibility of the conductor, who already has enough to do. Additionally, the events being monitored would be occurring at a location too far away for immediate action to be taken. The cameras may see a mid-train derailment as it happens, but what could be done about it?
I'm also not so sure that a small camera exists that could survive the conditions that exist in a railroad environment.
With the great advances in electronics I have an idea that could possibly enhance train operation and monitor the train. Have 2, 4 or possibly 6 portable cameras about the size of a Gopro camera mounted along the length of the train. They could be built with lockable mounts to quickly attach them to climbing rungs on the side of cars. The lenses could pan and tilt by remote control and be monitored live inside of the locomotive. This could detect unsecure loads, tresspassers, unwanted passengers, bad equipment or derailments. With todays technology I think this could be done without costing a lot. I'm not going to develop this idea so if anyone who reads this is interested, go for it.
Modeling the "Fargo Area Rapid Transit" in O scale 3 rail.
Our community is FREE to join. To participate you must either login or register for an account.