Link, but the Trib seems to need login now (and yes I registered so I have no clue if it is open)
Train leaves station and finds a car fouling a crossing. Can’t stop, boom, dead. That’s the jist of it.
Here’s my question: Is there a technology that can detect a fouled crossing at the time of signal activation and transmit the information to the cab, the dispatcher, or maybe a notification device set back from the crossing? I know it would be an extra cost to the railroad, but we don’t seem to be running out of people getting caught on the tracks. It is almost a daily occurance on Metra it seems some weeks.
Yep. Cameras at every crossing that transmit video pics to a video monitor in the locomotive/cab car. As soon as a train crosses a grade crossing, the next upcoming crossing is displayed.
It’s simple. I’ve seen something similar in Europe.
There are any number of possible methods of determining something is on the crossing - lasers, ultrasonics, mass detectors (like they use to trip traffic signals), CCTV, and who knows what else. Cost is obviously a factor, but we’ll make that secondary.
Any detector is going to have failings, so it’s a matter of which method will provide the most reliable detection of the desired objects.
A grid of laser beams would detect almost anything in it’s path, however it could easily be defeated with a piece of electrical tape (local residents think the train goes to fast?) or variations in visibility (fog, snow, etc). Vandalism could render it unusable, too. This might only result in the train passing through the crossing at restricted speed, but that could be a problem on a busy line.
Ultrasonics would face many of the same issues as laser beams, plus wind, which could literally blow the soundwaves off track.
Mass detectors would probably only be effective for vehicles, meaning a pedestrian in the gauge would not be detected.
Someone would have to monitor the CCTV, unless you wanted to equip all of the locomotives (and cab cars if applicable) with the ability to pick up a transmitted image from the camera.
At a high level, yes, technology exists. At a practical or economic level, no.
A bulletproof solution for which technology has existed for more than a century is to make each grade crossing into a manual interlocking, with an operator standing there in the tower 24/7 who doesn’t clear the home signal for the interlocking until the gates have come down and the crossing is clear of traffic. Barrier gates would prevent most but not all rubber-tire vehicles from crashing through and fouling the track. Track capacity would decline markedly and cost of operation would increase markedly. Roadway capacity would also decline markedly because the gates would have to come down much earlier. A typical grade-crossing warning time is 30 seconds (for single track, no traffic-light intertie – it increases with those) which is not sufficient braking distance for a train moving 79 mph. With any sort of an intertie between grade-crossings and wayside signals the gate-down time prior to arrival of train will advance to 3-10 minutes depending upon track speeds and descending grades. The gate-down times will start to overlap. In busy commuter territory the gates might go down for good at about 6 am and won’t be coming back up until 11 pm.
A cheaper, even more bulletproof solution is to grade-separate.
I think what you’re wondering is whether there is some type of electronic apparatus available today that can detect the presence of a car on the track and intertie with the wayside signaling system or to a locomotive on-board train-control system to either notify the train to brake to a stop, or to automatically brake the train to a stop. This can be done – although not totally reliably – with clear-out loops in the pavement that count rubber-tired vehicles entering and exiting and won’t clear the crossing for the train until the delta between the numbers is zero. They are used for quad-gate Quiet Zone crossings to cont
You’re seeing a solution for different parameters. First, grade crossings are much fewer in number in Europe; spacing is much wider and for trains at high speeds grade crossings are rare or nonexistant. This solution substantially reduces highway capacity. Trains in Europe are much lighter (fewer tons per operative brake) and thus have shorter braking distances. To implement this solution in the U.S. on a broad scale would result in enormous traffic gridlock and enormous loss of track capacity.
Another drawback of this solution is it requires an onboard electronic system. Per locomotive it might only be $10,000 but over a fleet of, say, 20,000 locomotives used in free-running through freight service it adds up to $200 million not including substantial O&M costs. Per grade crossing it would probably be at least $20,000 not including substantial O&M costs, and assuming you already have bandwidth available. If not you can ring up tabs of $1 million per mile adding the bandwidth, and as we’ve learned from some prototype installations in the U.S., getting “seven nines” reliability out of continuous RF transmission from wayside devices has proven virtually impossible. I know these numbers might seem absurd but manufacture and field-installation of safety-critical electronic hardware that has to run 24/7/365 regardless of temperature, climate, vibration, casual vandal attack, birds, bugs, lightning, grease, dirt, wind, costs a lot of money.
This solution does have site-specific usefulness and there are already applications of it in the U.S. on a site-specific basis, notably for industrial railroads. I
Little off topic, but I found it interesting that on the same page where this accident is being reported, there is a link to “Pedestrian hit by semi [truck].” Of course, with regards to this latter story, there will be no discussions along the lines of “Should the crosswalk have been protected by gates?” or “Why didn’t they build an overpass instead of a crosswalk?” or “Is there a way to have trucks automatically stop when there is a pedestrian in the crossing?” Just my [2c].
lol or they could just install scales under the crossing… like if a weight more than 100lb is on the crossing for more than somany seconds it throws an alarm to whoever… [sigh]
I like the camera at every crossing idea though with it switching to the next crossing on a monitor in the cab when they pass the current one, that would help alot i think
The best detection devise is “Do not do stupid things”.
Do not stop on the tracks. When the cars ahead of you stop. Do not move up to their bumpers to stop.
When I see cars stopped on the tracks I stop way before the crossings. If the gate start to come down. You can be sure they will want you to back up into the guy behind you so that they can back clear of the tracks. They will also blame you for not letting them clear the crossing.
IT is always someone’s elses falt that stupid things happen to them.
Lets say a freight train takes about a mile to stop at 60 mph. At 60 mph it takes a minute to go 1 mile. And you have to make a decision BEFORE you pass the mile mark. So you have to decide over a minute before you get to a crossing whether or not to stop. That’s a long time. The train starts blowing for the crossing about 1/4 to 1/2 mile away. So way before the train even has to blow for the crossing it has to begin braking.
If you shorten the time, you end up not being able to stop regardless of whether the crossing is blocked.
The best solutions are to retire the crossing, put in barrier gates or an overpass.
Mr. Husman’s point is a critical one, and is being missed by the other comments. There are plenty of possible ways to detect the presence of vehicles on railroad crossings. And they aren’t all pie-in-the sky either - there are devices currently in use on the US rail system. For example, in 4-quad gate installations, detection loops are typically used to prevent the exit gates from coming down if there are vehicles on the crossing, so vehicles aren’t trapped between the gates. I’m not aware that any of the existing detection loop devices transmit anything to a train, since that’s not the reason they are used. But I imagine something could be developed along the line of the automated messages sent by detectors
The issue isn’t detecting the vehicles in a crossing. Rather, the issue is whether it’s practical to have a system where vehicles are detected in time for a train to stop before reaching the crossing (without putting the train in emergency, which can create its own safety hazards), and which does not create false alarms by being triggered by normal, compliant vehicle traffic. The answer is “no”. This kind of system wouldn’t work at all on passive crossings (except on low speed lines), as the vehicle traffic is unlikely to stop until several seconds after the train begins whistling. A train moving 45 MPH or faster will begin whistling no more than a quarter mile away (FRA rules prohibit whistling further away, on the grounds that it is ineffective). Slower trains would begin whistling 15-20 seconds from the crossing (again mandated by FRA rules). This will typically be too short a time/distance for a train to stop before reaching the crossing. Theoretically, where active grade crossing signals are in use, it may be possible for the signals to activate far enough in advance of a train to permit responsive action for vehicles disregarding the signal
Actually, not only is it a critical point, it is the only relevant one. All other thoughts (other than the idea of eliminating the crossing) are at best impractical, at worst they are ridiculous.
[quote user=“Falcon48”]
There are plenty of possible ways to detect the presence of vehicles on railroad crossings. And they aren’t all pie-in-the sky either - there are devices currently in use on the US rail system. For example, in 4-quad gate installations, detection loops are typically used to prevent the exit gates from coming down if there are vehicles on the crossing, so vehicles aren’t trapped between the gates. I’m not aware that any of the existing detection loop devices transmit anything to a train, since that’s not the reason they are used. But I imagine something could be developed along the line of the automated messages sent by detectors
The issue isn’t detecting the vehicles in a crossing. Rather, the issue is whether it’s practical to have a system where vehicles are detected in time for a train to stop before reaching the crossing (without putting the train in emergency, which can create its own safety hazards), and which does not create false alarms by being triggered by normal, compliant vehicle traffic. The answer is “no”. This kind of system wouldn’t work at all on passive crossings (except on low speed lines), as the vehicle traffic is unlikely to stop until several seconds after the train begins whistling. A train moving 45 MPH or faster will begin whistling no more than a quarter mile away (FRA rules prohibit whistling further away, on the grounds that it is ineffective). Slower trains would begin whistling 15-20 seconds from the crossing (again mandated by FRA rules). This will typically be too short