mindthegates.uk - An Interesting Site on Level Crossings

[liamf656]

Not sure if there has been a discussion on this previously, but it is definitely worth sharing this website with the forum

You can view every level crossing in the UK, see when the barriers are forecasted to be down, how long for, which services pass, and also when the busiest times of the day are!!

Mind the Gates — UK level crossing live status & forecast

Know before you go. Live status and closure forecasts for all UK level crossings.

mindthegates.uk

 

[Farigiraf]

Definitely vibe coded. Layout looks very similar to some of the other AI-assisted projects posted this week on the forum - is it just the one person making all of these, or is this just the AI's style?

 

[bluenoxid]

Just looking at the Signal Box controlled crossing at Harrogate. Is it the walking route from the signal box?

 

[pokemonsuper9]

Apparently Navigation Road's crossing only closes 14 times a day, with a total closed time of 1 hour per day.

I imagine it's just pulling Network Rail's data for berth occupation and has no idea where the trams are.

 

[Bald Rick]

That site needs a much clearer safety disclaimer. I wouldn’t want to be whoever is responsible for that site if someone ends up misusing a crossing due to a data issue.

 

[Steddenm]

It does actually have a warning on there:

Not a safety device. Always follow the actual signals, barriers, and instructions at the crossing itself. Predictions on this site can be wrong — signalling failures, gatekeeper-held manual gates, and barriers held down for road events (school crossings, road-works convoys) can all produce closures this site won't show, and the live timeline can briefly gap during site maintenance. Never rely on this site to decide whether it's safe to cross.

 

[zwk500]

It does actually have a warning on there:

Buried in the 'about' section.

It feels like a good idea for a coding project that just didn't think through all the implications before uploading publicly with live data. IMO it'd be far, far better if the owner of the site revised it to present historical data only (even data from the previous day).

 

[liamf656]

Apparently Navigation Road's crossing only closes 14 times a day, with a total closed time of 1h per day.
I imagine it's just pulling Network Rail's data for berth occupation and has no idea where the trams are.

The same applies at David Lane in Bulwell. It was the first one I checked due to it's high usage with the NET Trams but it only shows the Robin Hood Line trains. It will be good if eventually these ones show up the tram data too

 

[DelW]

Looking at the site on Edge browser / Windows laptop, the safety warning is in bold text on the front page.

 

[eastwestdivide]

It seems to have Brinsworth Street LC, on the Holmes Chord in Rotherham, as open for a two-hour stretch right now, which is entirely wrong.

Later edit: and now it’s showing as closing for a Tinsley-bound freight which wouldn’t pass the crossing (wrong line), so the location must be wrong in their system. That’s followed by an open period for over an hour, during which there will be a number of passenger trains.

 

[liamf656]

It seems to have Brinsworth Street LC, on the Holmes Chord in Rotherham, as open for a two-hour stretch right now, which is entirely wrong.

Looks like for some reason it's failed to bring up any passenger services for that crossing. The nearby Holmes Junction crossing is bringing up more accurate information

 

[edwin_m]

The same applies at David Lane in Bulwell. It was the first one I checked due to it's high usage with the NET Trams but it only shows the Robin Hood Line trains. It will be good if eventually these ones show up the tram data too

They are sort of right for that one. The wigwags and barriers only operate for trains - if there's only a tram the traffic is just stopped by normal traffic signals.

 

[eastwestdivide]

Looks like for some reason it's failed to bring up any passenger services for that crossing. The nearby Holmes Junction crossing is bringing up more accurate information

Yes, I think it's erroneously placed the crossing on the line towards Tinsley, not the one towards Meadowhall. But if that's the case, why didn't it pick up the tram-trains that use that section, and which show up in the data feeds like RTT?

To quote the site's front page, "predictions may be wrong".

 

[Lucan]

I wouldn’t want to be whoever is responsible for that site if someone ends up misusing a crossing due to a data issue.

Are you seriously saying that people would crash through closed barriers just because that website predicted they would be open at that time?

 

[Bald Rick]

Are you seriously saying that people would crash through closed barriers just because that website predicted they would be open at that time?

It is not unknown for pedestrians or cyclists to jump barriers, nor for motorists to weave around half barriers through impatience, or in the belief that they have enough time to cross before a train is coming, or that the barriers are in some way broken.

When I say ‘not unknown’, I mean it happens every single day, multiple times. If people have a website that erroneously suggests that the crossing should be open when it isn’t, I suggest that a small proportion of some people, with less patience than average, and with less intelligence than average, may be more likely to take that course of action, yes.

 

[AlterEgo]

Are you seriously saying that people would crash through closed barriers just because that website predicted they would be open at that time?

One of the biggest causes of level crossing misuse is the belief a train is not due to actually come. People swerve around barriers often or try to jump/beat them when they are closing.

 

[High Dyke]

Interesting to see some of the data shown. As an example I tried a number of crossings I am familiar with. The ECML crossings I can understand long closures. As an aside I tried one location I work, with 2tph. However the crossing is effectively a restricted byway, so we rarely open the gates between trains. I note the disclaimer.

 

[Lucan]

It is not unknown for pedestrians or cyclists to jump barriers, nor for motorists to weave around half barriers through impatience

One of the biggest causes of level crossing misuse is the belief a train is not due to actually come. People swerve around barriers often or try to jump/beat them when they are closing.

Yes, of course, on this forum we all know that. But my point was : are people going to do that just because they consulted that website?

Those people will do it because they saw the gates closing in front of them and don't want the wait, just like they have always done. I very much doubt they will stop at the wig-wags and bring up that website on their smartphone to help them decide whether to make a dash for it.

 

[AlterEgo]

Yes, of course, on this forum we all know that. But my point was : are people going to do that just because they consulted that website? Those people will do it because they saw the gates closing in front of them and don't want the wait, just like they have always done. I very much doubt they will stop at the wig-wags and bring up that website on their smartphone to help them decide whether to make a dash for it.

Nobody is suggesting that, but the presence of a site which suggests trains are not coming and specifically predicts the barriers should be up adds a significant risk, especially as it is quite obviously wrong a lot of the time.

 

[Bald Rick]

I very much doubt they will stop at the wig-wags and bring up that website on their smartphone to help them decide whether to make a dash for it.

But someoone might. Especially someone who doesn’t understand the railway, and has been pointed to a website that does show when the barriers should be open. It only takes one person to do it, and end up in front of a train, and then the police will likely get involved…

 

[matt1024]

Hi all, site builder here. Thanks liamf656 for posting it, and everyone who had a look. Still a work in progress, this thread has been really helpful. Some fixes (hopefully) since yesterday:


Brinsworth Street (eastwestdivide) - you had it right, it was attached to the Tinsley line instead of the Holmes Chord. Rebound, hopefully showing the correct line now. Checking other crossings near junctions for the same mistake.

Trams (pokemonsuper9, liamf656, edwin_m) - correct, they were invisible. I can’t find live feeds for most tram lines, instead those crossings now use the published tram timetables, labelled as such.

Harrogate (bluenoxid) — NR's census has it as a staff barrow crossing. Removed, plus about forty similar staff-only ones the same check found.

@High Dyke - fair point, "open" there really means "no train due". If you wouldn’t mind tell me the crossing in question, I can update the model.

Warnings (Bald Rick, zwk500) - fair concern. Every page now says the status is predicted, with the warning right under it rather than just the footer. I'll keep tightening this.

 

[SpannerWeilder]

Rainham (Kent) doesn't take into account that the Thameslink services start and end in the bay platform and do not trouble the level crossing.

 

[zwk500]

Hi all, site builder here. Thanks liamf656 for posting it, and everyone who had a look. Still a work in progress, this thread has been really helpful. Some fixes (hopefully) since yesterday:

Hi Matt, it's an interesting site! What was/were the intended use(s) of the website was when building it?

 

[Meole]

Interesting to see which manned level crossings have very little activity.

 

[matchmaker]

Looking at my local crossing (Cornton 1) it shows Edinburgh to Dunblane trains passing through Causewayhead Junction. Which isn't on that line...(it's on the Alloa line)

 

[styles]

Must've been the same software development outlet as the new ScotRail app and the Emirates Skywards Rail website.

Causing more problems than it solves.

 

[PhilipRdNGauge]

Hi all, site builder here. Thanks liamf656 for posting it, and everyone who had a look. Still a work in progress, this thread has been really helpful. Some fixes (hopefully) since yesterday:


Brinsworth Street (eastwestdivide) - you had it right, it was attached to the Tinsley line instead of the Holmes Chord. Rebound, hopefully showing the correct line now. Checking other crossings near junctions for the same mistake.

Trams (pokemonsuper9, liamf656, edwin_m) - correct, they were invisible. I can’t find live feeds for most tram lines, instead those crossings now use the published tram timetables, labelled as such.

Harrogate (bluenoxid) — NR's census has it as a staff barrow crossing. Removed, plus about forty similar staff-only ones the same check found.

@High Dyke - fair point, "open" there really means "no train due". If you wouldn’t mind tell me the crossing in question, I can update the model.

Warnings (Bald Rick, zwk500) - fair concern. Every page now says the status is predicted, with the warning right under it rather than just the footer. I'll keep tightening this.

Hi

As a prediction tool, I can see it's usefulness. I have in the past used RealTimeTrains to predict the state of Thorpe Lane Crossing when coming back from Egham to Staines, with an impatient driver. If it's likely to be open when we get there it's a shorter route than the A30, but if not you're better off going via Egham town, even if that crossing is closed as it tends to clear quicker. Whatever logic you've developed would be useful attached to things like Google Maps etc, to add suitable delay minutes to a route which takes you over a level crossing if it's likely to be closed when you get there & so giving a better guess at which will be the fastest route.

Keep up the good work

 

[kieron]

Looking at Balderton in Cheshire, I see a few problems.

1. Like with some other places, it includes trains which don't go through the level crossing. Here some of the trains it lists get no closer than Whitchurch, 15 miles away.

2. It doesn't take VSTP changes into account. The Wrexham–Crewe trains were cut back today to just Chester–Crewe (so don't use the crossing), but it lists them as if they did.

3. In the "Trains in window" list, the "from xxx" section seems arbitrary. Here it lists southbound trains as "from Chester" and northbound ones as "from Chester Middle Yard", both of which are north of the level crossing.

 

[Islineclear3_1]

3 min closure for Newhaven Town...? More like 7-8 minutes for trains heading in the direction of Seaford

 

[matt1024]

Hi Matt, it's an interesting site! What was/were the intended use(s) of the website was when building it?

I built it as there's a local crossing I keep getting stuck at, I wanted to know before I left whether to go over the crossing, take the long way round or wait for a gap. Initially it just covered this one crossing, but I thought other people are probably in the same predicament with other crossings and I thought it would be a fun challenge to try and cover every crossing. Exactly like:

Hi

As a prediction tool, I can see it's usefulness. I have in the past used RealTimeTrains to predict the state of Thorpe Lane Crossing when coming back from Egham to Staines, with an impatient driver. If it's likely to be open when we get there it's a shorter route than the A30, but if not you're better off going via Egham town, even if that crossing is closed as it tends to clear quicker. Whatever logic you've developed would be useful attached to things like Google Maps etc, to add suitable delay minutes to a route which takes you over a level crossing if it's likely to be closed when you get there & so giving a better guess at which will be the fastest route.

Keep up the good work

Thanks Philip, appreciate this. I have been thinking about integrating into maps, it's a fair amount of work, but probably doable. First I want to iron out any issues and make the predictions as good as they can be (I know it can never be 100% accurate).

Looking at Balderton in Cheshire, I see a few problems.

1. Like with some other places, it includes trains which don't go through the level crossing. Here some of the trains it lists get no closer than Whitchurch, 15 miles away.

2. It doesn't take VSTP changes into account. The Wrexham–Crewe trains were cut back today to just Chester–Crewe (so don't use the crossing), but it lists them as if they did.

3. In the "Trains in window" list, the "from xxx" section seems arbitrary. Here it lists southbound trains as "from Chester" and northbound ones as "from Chester Middle Yard", both of which are north of the level crossing.

Thanks Kieron, this was very helpful. #1 should be fixed now, the railway network graph we source is missing co-ordinates for some of the nodes, in these cases we have to try and deduce it, we got it wrong here. (This fix also applied to ~25 other crossings) #2 hopefully also fixed now, getting VSTP right has been an ongoing challenge, in this case a same-day re-plan arrives over VSTP as a replacement schedule, not a cancellation as I first assumed. (This fix also applied to ~150 other crossings).

For your third point, and yours @matchmaker , I'm going to think about how to best handle this as the fix isn't as simple and as it might seem. Originally I labelled trains as "up" and "down", but this doesn't mean much to people not familiar with the terminology, so I changed it to essentially "find the next identifiable location along the line", which turns out to be quite hard to define programmatically in a way that is correct and meaningful for every crossing.

Rainham (Kent) doesn't take into account that the Thameslink services start and end in the bay platform and do not trouble the level crossing.

Should be fixed now, similar issue to #1 above (This fix applied to ~40 other crossings).

3 min closure for Newhaven Town...? More like 7-8 minutes for trains heading in the direction of Seaford

Thanks, sounds plausible. We do support line and crossing specific timings, but I don't want to adjust timings off memory (mine or anyone else's) they need calibrating. I have published a calibrator for exactly this purpose. If you're ever stood at the crossing for some time, any help fine tuning the timings is very much appreciated.