How to use Mapillary for watching images close to railroad crossings in Europe?

Imagery, data, and integrations
1

Hi, please let me know how to use Mapillary for watching images close to railroad crossings in Europe

I would like to rate railroad crossings in Europe basing on risk opinions by peers and on statistics of accidents.

Mapillary offers a nice interface integrating OpenStreetMap tiles with street level images of road and railroad crossings.

I wrote a script to mark railroad crossings in OpenStreetMap with a raised icon.
If I click an icon I get id tag, name of a specific railroad crossing and its rating.
I can click a link to Mapillary to open image of a railroad crossing.

Script worked fine yesterday but today failed to connect to Mapillary or image is black.

I would like students to watch slideshow of images from both sides of the specified railroad crossing
to let us discuss risks involved, make opinions and assign risk label:
High risk
Low risk
No risk

Please let me know if Mapillary can support such educational activity.

European Commission for the Investigation of Accidents at Railroad Crossings

EU Road Safety Expert

2

hello, sounds like a good initiative. If the script worked before, could it be broken by (recent?) API changes?

3

Thank you. I can update my script, but the key obstacle is the large number of railroad crossings—over 200,000 in Europe and over 370,000 in the U.S. —which is why I’m looking for railroad crossing experts, railroad crossing Watchers, and citizen scientists specializing in railroad crossings, so that we can identify the most dangerous railroad crossings in the world (Europe, the U.S., Canada, Australia…).

Since we live in a world where time is of the essence, I would like to write scripts that automatically link every railroad crossing accident reported on social media (Twitter, Facebook, Instagram…) with street-level imagery, so that we can publicly discuss each case, analyze the causes of such accidents, and rate a given crossing as high-risk, very dangerous, and one that should be avoided if possible. By working hard, we can reduce the number of railroad crossing accidents worldwide by 50% this year. If you live near a railroad crossing, you can serve as an RC Watcher, submiting detailed photos of that crossing, and add public comments that will be saved under Comments tag for that specific crossing. I need to write an AI5.0 script that will automatically generate a link with geolocation data (longitude and latitude) to street-level and satellite imaginery as a comment to an RC accident report posted on social media. I can obtain accident statistics by RC name for Australia. I’m looking for similar data for Europe, the U.S., Canada, and other regions of the world.

4

Are not track owners and operators usually obliged to record and report accident statistics on railroad crossings to some national transportation oversight authority? Furthermore, track operators usually also have to report and classify railroad crossings to safety standard levels to oversight authorities too. Most of the time such data should be in public domain or retrievable through freedom of information acts. So, I do not exactly understand why you de facto try to double existing effort that most often has already been payed for by taxes and is backed by existing real world data than some felt safety score. In my understanding, if you want to improve the safety of railroad crossings in general it should probably be best to constantly lobby decision makers to raise each railroad crossing to the maximum safety standard level. But, maybe I am missing something?

1 Like
6

Ok, you are right

You can visit
Highway-Rail Grade Crossing Safety | FRA

But what I mean is how to access a single safety certificate for the given railroad crossing to make car driver to never risk his/her life, crossing it.

40 years ago, I drove to a golf course to visit some friends, and then I continued on to see my family. I came to a large railroad crossing closed off by a heavy barrier; there were four railroad tracks at the crossing, and I had about 50 meters to drive before reaching the exit barrier.
The barrier was open, standing upright, with no red flashing lights, and as soon as I drove past the entrance barrier, it began to close—and the hair on the back of my neck stood on end, just like it has for millions of drivers—wondering whether I should park my car across the tracks or if I’d make it across before the exit barrier closed.

I made it and I’m alive, but thousands of drivers were trapped between the closed barriers and either died or had their cars smashed to pieces.

Today we have the internet, OpenStreetMap, and Mapillary, but as I approach the crossing and drive onto an open one, I still don’t know if I’ll get trapped—if the entry and exit barriers will close—or how many seconds until a train arrives, at what speed, and from which direction.

Railroad supervisors train drivers to ram the barrier with their car and flee the crossing in such a situation.
But they point to a closed entry barrier and do nothing to ensure that the exit barrier remains open—as in my case—so that the driver can make it across the crossing.

Hence my interest in Mapillary—to gain this insight by analyzing photos of the crossing, street-level views, and satellite imagery.

1 Like
7

Please watch this video from railway crossing to learn how Mapillary can save thousands of car drivers, generating pre-warning to drivers approaching the same railroad crossing in a future.
(Car driver trapped between closed gates Syndrome)

1 Like
8

The anecdote you are telling is either a regulatory or implementation issue. Usually, multiple track railroad crossings have like a delay time factor for each track before the barrier on the other side closes. This is similar to how traffic signals have to be set up before they go red at intersections for roads with different speed limits. Although, it is true that some jurisdictions mindlessly ignore physics, like in the UK. Furthermore, the barriers have to be sufficiently easy breakable from the inside with a vehicle.

So, in your case either the regulation did not mandate such a delay time factor or the delay time factor was not implemented correctly on that particular railroad crossing. In my understanding, the natural course of action should be to either lobby for such a delay time factor in the regulation, extend it, or report the lack of implementation to the authorities. Written reports or complaints are indeed tedious to create but they are also hard to ignore for the bureaucracy. In my experience, bureaucracy does indeed move slowly. However, often once it gets moving it moves steadily. Usually, you will get a negative response initially because bureaucrats love to show off their imagined superiority over citizens (they never make mistakes, of course) and also go the path of least resistance because they do not like change since they have been primarily taught to keep things running rather than changing or improving anything. A negative response is no reason to give up. Analyze the response and complain that they did not actually focus on your request. Yes, it is boring and tedious but it often works.

Anyhow, I do understand your good will to improve safety at railroad crossings but I still fail to see how a felt safety score or buzz on social media is going to change or improve anything because neither of these have any impact on how bureaucratic systems or democracies work. Besides, do you really believe you will find a critical mass of people that score, report, or discuss anything in this domain? Even if, there are still going to have to be people who will push the cogs in the bureaucracy, regulatory, and legislative bodies. There are dozens of accidents at railroad crossings daily around the world. None of them spike any public outcry unless mothers with babies die in a fire. It is a sad but honest truth.

9

My Dear Friend,

thank you for your kind reply and comments.

But life is for real:

"At Least 64 Dead in Spain’s Toledo Train Collision: The Second High-Speed Disaster in Nine Months

"a spokesperson for the Spanish Ministry of Transport explained that, according to initial reconstructions, the AVE train may have hit the Iryo train, causing the derailment of two carriages and their fall into an embankment. According to what was reported by El PaĂ­s, the most accredited explanation would be human error, since following the malfunction of the safety system around eight trains stopped in the area.

The safety system is automated, but in the event of a breakdown it switches to manual mode. At this point the Adif technicians intervened, ordering the stopped convoys to gradually restart, but the Iryo train was unable to do so and remained stopped. According to initial rumours, an operator at the control center at Atocha station in Madrid mistakenly gave the order to the Renfe train to restart despite the Iryo blocking the tracks. The Renfe train then collided with the Iryo from behind at an as yet undetermined speed, causing the accident.
"

So I develop some safety standards to let us, individual car drivers to avoid any risk, when crossing railroad crossings in Europe or anywhere else in the world.

Please read the information about today’s train accident in Spain to understand why we shouldn’t waste time dealing with railway bureaucracy.

They’re still living in their own world, 100 years behind the real world, which is full of innovative technologies.

Trains in Europe still do not have pre-installed GPS trackers, and the operator does not know their current location on the map; that is why they failed to react and prevent the collision, which resulted in nearly 100 casualties.

A similar disaster years ago brought an end to the MagLev magnetic levitation rail experiment in Germany.

10

No matter whether you like it or not, railroad oversight authorities and regulatory bodies will continue to exist and they will continue to have final say in what safety standards are required and how they have to be implemented by rail operators. All I am saying is that one should try to be as effective as possible in every domain. Here, it means that you have to find a lever that pushes change in the bureaucracy because this is how railroads work. Do not waste your time and effort on stuff that is not going to improve rail safety effectively. Focus on tangible and measurable improvements, not some soul-soothing good deeds.

No, Transrapid development ceased because the German federal government ended taxpayer funding, when they finally came to their senses and realized that there are no and never will be enough serious customers for the system because it requires building completely new tracks and stupendously expensive infrastructure. The dual steel rail track system may not have won technologically but it has already won commercially. The pseudo‑commercial Shanghai maglev train is and always was just a political project meant as a bargaining chip to unlock access to the Chinese market and labor force for the German economy. The disaster you are referring to happened on the only existing experimental test track loop in Germany, where a Transrapid passenger car with tourists crashed into a maintenance car on the closed loop track. The track operators did not check the work schedule. Both accidents have nothing to do with railroad crossings.

11

My Dear Friend,

thank you for your comments.

But railway crossings in Europe are controlled by the same staff, who was responsible for 2 high-speed train crashes in Spain in 2026.

Please read the transcript (from X):

"
Closing the circle further on the Urda accident.

At 8:30 a.m., a computer failure knocked out the LZB, the system that monitors train speeds on the Madrid-Sevilla line.

Six trains stopped on their own.

Up to that point, everything worked as it should.

What came next fits into seventeen minutes.

8:36. Adif’s control center in Atocha orders to resume travel “normally.”

Without LZB, the protocol demands the opposite: proceed by sight, 40 km/h maximum, and eyes on the track.

The order wasn’t just for the AVE.

Other trains received the same mistaken instruction, and one of their drivers corrected the controller: no, by sight.

The error went unnoticed.

Simply put, it wasn’t corrected for everyone.

8:41. The Iryo contacts the control center and receives the same instruction: resume normally.

8:48. The Iryo driver calls again. The train won’t let her restart the system. It’s stopped.

Meanwhile, the AVE is advancing at over 180 km/h on average in the minute before impact.

8:53. The AVE rear-ends the Iryo at the entrance to Urda station.

Sixty-four dead up to this moment.

It’s what’s hardest to understand.

On the controller’s screen, the Iryo appeared stopped.

The AVE was approaching it at a speed the protocol didn’t allow.

There were seventeen minutes between the order and the crash, and at least one warning five minutes earlier that the train ahead couldn’t move.

And not a single braking order.

Nothing makes sense.

Or maybe it does, and that’s what’s worrying.

An isolated error is just that—isolated.

But this error repeated with several trains, one driver spotted it and corrected it, and yet everything continued for seventeen minutes with all the information on screen and five minutes before the collision the struck train says it’s stopped.

That’s no longer a slip-up.

It’s a system where no one, or nothing, had the capacity to understand what was about to happen and stop it in time.

That’s the question the investigation will have to answer.

Not who messed up at 8:36—that seems clear—but why no one could correct it at 8:50.

A buzz, a siren, something in the control room.

I am a small fish, so I develop
Pre-voice warning in car navigation:

You are approaching railway crossing
Status: High risk

Railway officers rejected installation of GPS tracking in trains by default
Railway officers rejected installation of a monitor in a train to let a driver to watch live image from camera/s installed over the railway crossing
Railway officers don’t install inductive loop under the railway crossing to detect a car as an obstacle
Railway officers showed no interest in installing a countdown timer before the railroad crossing to inform drivers how long until a train arrives, from which direction, and at what speed

I am a small fish and must live with the thought, that when I cross a dangerous railroad crossing, I’m always risking my life

12

**

Radio 5**

@radio5_rne

·

6h

Translated from Spanish

đź”´nio Tobal, Head of Safety in Railway Operations for UGT at ADIF

🗣️"The colleague was managing the incident with eight trains at a standstill. There are constant phone calls. That involves a series of mental processes, under pressure and stress, a professional, no matter how well-prepared they are, can have that lapse."

Just attached to make you aware of the risks, while approaching railway crossing in Europe.

You always risk your life as a car driver, cyclist or a pedestrian, since safety standards are not met

"Antonio Tobal, Head of Safety in Railway Operations for UGT at ADIF

13

The biggest problem at level crossings is road users. Even when the warning system is working, there are always road users who think they can still drive or walk across. There are also many road users who blindly follow the vehicle in front of them. If that vehicle then comes to a halt, you’re left stationary in the middle of the level crossing. Despite all the warning systems, they fail to achieve their purpose if people do not change their behaviour. The highest level of safety is achieved by removing the level crossing altogether. This is also what is happening in the Netherlands.

2 Likes
14

Exactly, the best railroad crossing is no crossing. Unfortunately, this is not always physically possible to retrofit. However, what citizens can do is to push and lobby for removing railroad crossings everywhere it is possible. Ironically, no crossing does not only improve safety but also everybody’s balance sheet at the end of the day. But hey, why bother bureaucrats who are fed taxpayer money? :person_shrugging:

Yes, the :netherlands: is light years ahead of pretty much everybody in road safety design and transportation infrastructure architecture. Afaik, :japan: has also had a grade separation and level crossing elimination program for a long time too.

1 Like
15

Please read and watch

"

Espartano Estoico

@7vecesEspartano

·

A train passes through Asua, Erandio, Biscay with the level crossing barrier raised. And the barrier lowers after the train has passed, there are no more railway accidents in Spain because God doesn’t want it that way.