Random thoughts on measuring Red Light Jumping

mo_bhicycle · 23-04-2019 11:00pm #1

I'm probably opening a can of worms here, but I'm curious to hear what others think.
It occurred to me that looking at absolute numbers of cars & bikes that RLJ is a bit skewed as both groups don't have the same opportunity to RLJ.

Take the picture below for instance, the front two rows of cyclists could conceivably RLJ if they wished, while the lone car has no opportunity.

In the picture below the front 3 cars could conceivably have an opportunity to RLJ, while those behind cannot unless the ones in front RLJ.

I wonder if we looked at RLJers per cars/bikes with an opportunity would our perception of problem junctions change.

What I'm talking about is something along the following lines:
rlj_opp = #rljers/#cars_or_bikes_with_rlj_opportunity

ted1 · 23-04-2019 11:10pm

Cars v bikes ? Think I’ll grab the popcorn.

Wishbone Ash · 23-04-2019 11:16pm

It's pointed out ad nauseam here that motor vehicle drivers would jump red lights a lot more frequently if they could but once the vehicle in front of them stops they physically can't.

spurious · 23-04-2019 11:27pm

Wishbone Ash wrote: »

It's pointed out ad nauseam here that motor vehicle drivers would jump red lights a lot more frequently if they could but once the vehicle in front of them stops they physically can't.

It's only a matter of time before one of the the numerous cyclist RLJs at this and similar junctions smacks into an emerging car and then they will be in the 'physically can't' category.
https://goo.gl/maps/8oLJQHzw7cfPuLHr6
Frightening how many of them do it, even flying through narrowly missing kids despite the lollipop folk being there during school term times.

magicbastarder · 23-04-2019 11:41pm

red light 'jumping' seems like a relatively innocuous phrase. we need a better phrase. red light flouncing?

work · 24-04-2019 1:28am

Without enforcement this will always be an issue. OP just points out measuring is complex.
I find the point above made about going through lights with traffic wardens on them being dangerous interesting. It is COMPLETELY unacceptable.
However wouldn't it be interesting if we took a few junctions and statistically analysed the jumpers, all traffic, to see the chances of causing a fatality. I am confident if you excluded the cyclist himself dying by light breaking the fatality risk would be VERY heavily in favour of motor vehicles. Would love someone to quantify this and it might put some anti- cyclist arguments to bed......but In could be wrong!

24-04-2019 1:43am

It comes down to the person, though. If I'm gonna jump red lights, then I'll do it whether im on a bike or in a car.

Fighting Tao · 24-04-2019 2:46am

Didn’t the RSA manage to measure this already and publish the results? They may have published methods too.

SmartinMartin · 24-04-2019 2:50am

Deleted User wrote: »

It comes down to the person, though. If I'm gonna jump red lights, then I'll do it whether im on a bike or in a car.

Bull****. You are identifiable in a car, not so on a bike.

Wishbone Ash · 24-04-2019 7:17am

Deleted User wrote: »

It comes down to the person, though. If I'm gonna jump red lights, then I'll do it whether im on a bike or in a car.

The point of the thread is that it's not possible in a car once there is another car stopped in front. Whereas on a bike it usually is possible to filter through and break the light. In other words, a lot more motorists would break red lights if they physically could.

Bigus · 24-04-2019 7:24am

Wishbone Ash wrote: »

The point of the thread is that it's not possible in a car once there is another car stopped in front. Whereas on a bike it usually is possible to filter through and break the light. In other words, a lot more motorists would break red lights if they physically could.

a lot more motorists would break red lights if they physically could, get away with it .

07Lapierre · 24-04-2019 7:47am

Traffic cameras are being installed at a few junctions in Dublin shortly. Be interesting to see the RLJ statistics. I suspect the number of cars RLJ’ing will drop significantly, while the number of cyclists RLJ’ing won’t change, as cyclists can’t be identified as easily.

At the same time I suspect the number of cyclists RLJ’ing will be quite small anyway.

LeinsterDub · 24-04-2019 8:04am

It's a shame we can't have Idaho stops and more zebra crossings here https://en.m.wikipedia.org/wiki/Idaho_stop

.G. · 24-04-2019 8:04am

Need to add electric scooters into the mix now, loads of them ignoring red lights too.

Achasanai · 24-04-2019 8:17am

SmartinMartin wrote: »

Bull****. You are identifiable in a car, not so on a bike.

They may be identifiable, but if nothing is done, it doesn't really matter. There's a junction on my way to work, and I see at least one car breaking a red every day. And that's a straight red. An amber light has become a sign for most cars to speed up rather than what used to be a red light to stop (unless unsafe to do so).

Alkers · 24-04-2019 8:48am

Deleted User wrote:

It comes down to the person, though. If I'm gonna jump red lights, then I'll do it whether im on a bike or in a car.

I would never intentionally break a red when driving (car, campervan, motorcycle). I do it all the time on my pushbike

seamus · 24-04-2019 8:53am

Wishbone Ash wrote: »

The point of the thread is that it's not possible in a car once there is another car stopped in front. Whereas on a bike it usually is possible to filter through and break the light. In other words, a lot more motorists would break red lights if they physically could.

I was stopped at a light one day. There was nothing coming. I was on Amiens street heading southbound, stopped at the lights to allow traffic to come from behind custom house onto the quays.

I could easily have just gone, without incident. But it was a red light. I heard a little "Bip" behind me and car rolls up beside me, winds down the window and asked me why I'd stopped.

"The light's red"

"But...there's nothing coming?"

"Yeah, but the light's red"

"Can you get points on your licence?"

"No"

"Then why have you stopped?"

"Because the light's red"

He was absolutely bemused.

Given how much red light breaking I see every day and the fact that it's rarely just one ambler gambler, it's 3/4/5 of them at a time, I have zero doubt that red light jumping is most definitely an "all roads users" problem, and not one that bikes are even the worst offenders for.

Chips Lovell · 24-04-2019 9:12am

Wishbone Ash wrote: »

It's pointed out ad nauseam here that motor vehicle drivers would jump red lights a lot more frequently if they could but once the vehicle in front of them stops they physically can't.

That used to be the case, but more than once recently, I've seen someone drive around a car stopped at red in order the break the lights. The most idiotic example of this was at the Castleknock Gate of the Phoenix Park one morning, which of course is very narrow and risks getting the RLJ'er caught between the stopped car and oncoming traffic.

Wishbone Ash · 24-04-2019 9:26am

Chips Lovell wrote: »

That used to be the case, but more than once recently, I've seen someone drive around a car stopped at red in order the break the lights....

I accept that it happens and it happened to me on more than one occasion, but it's fairly rare. 99% of motorists will stop and wait when the vehicle in front has stopped.

smacl · 24-04-2019 9:31am

mo_bhicycle wrote: »

I'm probably opening a can of worms here, but I'm curious to hear what others think.
It occurred to me that looking at absolute numbers of cars & bikes that RLJ is a bit skewed as both groups don't have the same opportunity to RLJ.

Rather than guess, you'd be better off to measure and proceed on that basis. AI based object recognition tools such as tensorflow and YOLO are reasonably easy to setup and this point and work very well with video. If you set a video up record a junction when lights were red, you could automatically count cars and bikes breaking the lights and work from there.

https://www.youtube.com/watch?v=yQwfDxBMtXg

Chips Lovell · 24-04-2019 9:40am

Wishbone Ash wrote: »

I accept that it happens and it happened to me on more than one occasion, but it's fairly rare. 99% of motorists will stop and wait when the vehicle in front has stopped.

Definitely agree. I was more saying that I never used to see it happen and now I've seen it happen a few times.

Effects · 24-04-2019 12:39pm

spurious wrote: »

It's only a matter of time before one of the the numerous cyclist RLJs at this and similar junctions smacks into an emerging car and then they will be in the 'physically can't' category.
https://goo.gl/maps/8oLJQHzw7cfPuLHr6
Frightening how many of them do it, even flying through narrowly missing kids despite the lollipop folk being there during school term times.

I know that junction well. Cyclists do go through it on red, but usually at a fairly low speed. It's the drivers that routinely break it that worries me, as they travel faster and carry more inertia. Drivers seem to think they have a right to break lights cause they pay 'road' tax.

mo_bhicycle · 24-04-2019 1:19pm

smacl wrote: »

Rather than guess, you'd be better off to measure and proceed on that basis. AI based object recognition tools such as tensorflow and YOLO are reasonably easy to setup and this point and work very well with video. If you set a video up record a junction when lights were red, you could automatically count cars and bikes breaking the lights and work from there.

https://www.youtube.com/watch?v=yQwfDxBMtXg

I like that idea.
Thinking out loud here: I wonder if I got one of the magnetic Polaroid cameras (they’re quite good value atm) and put it on a light pole could I get good enough footage? It’d need to be high enough not to be stolen but not so high itd miss the action ðŸ˜›

rubadub · 24-04-2019 1:48pm

Simona1986 wrote: »

I would never intentionally break a red when driving (car, campervan, motorcycle). I do it all the time on my pushbike

same here. I would be more concerend with dangerous red light jumping, what the law actually set out to prevent. I have been in taxis very early in the morning who drove through completly empty junctions, I would not do that myself though.

I would always question "if the traffic lights were broken and you saw that person break the light in that manner would you have even have batted any eyelid?".

LeinsterDub wrote: »

It's a shame we can't have a Idaho stops

From what I see most gardai do sensibly view cyclists & pedestrians breaking lights to be like the Idaho stop.

I frequently break a light on my commute which is very near a garda station. Gardai used to be on duty in that area a lot and would often give me, what I took to be, a nod of approval as I broke the light, as they had sense and knew I was doing it for my own safety (and to aid the flow of traffic). Just as I have gotten waves from gardai while cycling on completely empty footpaths for my own safety and to let traffic pass.

It is extremely rare that I see cyclists on the cycletrack obeying this light, unless pedestrians are there, I break it myself but of course would slow or stop if cyclists are coming from the other side trying to join the cycletrack.
https://goo.gl/maps/BJDsjqXrXP5NBGUv7

Rechuchote · 25-04-2019 9:21am

magicbastarder wrote: »

red light 'jumping' seems like a relatively innocuous phrase. we need a better phrase. red light flouncing?

I always flounce through traffic lights.

mo_bhicycle (OP), people on bicycles and people in cars run red lights differently. On bicycles, people filter through a red light when there's nothing coming; and some ninjas don't look, the gobdaws. In cars, some drivers race through after the light has changed, at speed, so they won't have to wait.

It's a cliche of traffic study but it's true: cyclists go through red lights for safety; drivers go through red lights to save time.

RayCun · 25-04-2019 9:24am

Rechuchote wrote: »

mo_bhicycle (OP), people on bicycles and people in cars run red lights differently. On bicycles, people filter through a red light when there's nothing coming; and some ninjas don't look, the gobdaws. In cars, some drivers race through after the light has changed, at speed, so they won't have to wait.

It's a cliche of traffic study but it's true: cyclists go through red lights for safety; drivers go through red lights to save time.

I agree that people are breaking lights in different ways, but let's not pretend they are doing it for different reasons :pac:

People on bikes will break lights by slowing down to see if there is anything coming that will hit them, and if not going through. Or stopping when there is traffic going through that would hit them, and starting again when there isn't. (or sometimes going through and expecting everything else to stop for them, but that's not common)

People driving cars tend not to roll through lights (though I've seen it done), they usually break lights by going through shortly after it has turned red.

But both are doing it to save time.

Rechuchote · 25-04-2019 9:26am

RayCun wrote: »

I agree that people are breaking lights in different ways, but let's not pretend they are doing it for different reasons :pac:

They certainly are.

magicbastarder · 25-04-2019 9:27am

i posted this over on the commuting and transport forum earlier, was a slightly odd situation to see a driver berating a cyclist for *not* breaking a red:

magicbastarder wrote: »

at my favourite junction for RLJing - the junction of clontarf road and alfie byrne road - again.
i was on the bike, in the right turning lane, facing a red light for all traffic. the light went green for straight on only, and the three cars in front of me drove thorugh a solid red right turn light; i rolled up to the white line and the guy in the car behind me beeped at me and gestured for me to take the turn - i mouthed 'no' at him and he gave me an angry gesture, drove around me and went through the red.

Seth Brundle · 25-04-2019 9:28am

Rechuchote wrote: »

It's a cliche of traffic study but it's true: cyclists go through red lights for safety

Where are you getting that from?
There are occasions where a cyclist must break a light as it won't change unless a car approaches and triggers the sensor. I can't think of any safety reason though.

RayCun · 25-04-2019 9:32am

Rechuchote wrote: »

They certainly are.

It's like drivers arguing that they are forced to break lights because otherwise the car behind will run into them. :rolleyes:

The overwhelming majority of cyclists breaking lights are doing it to save time (and momentum!) not because it is safer.

smacl · 25-04-2019 9:37am

mo_bhicycle wrote: »

I like that idea.
Thinking out loud here: I wonder if I got one of the magnetic Polaroid cameras (they’re quite good value atm) and put it on a light pole could I get good enough footage? It’d need to be high enough not to be stolen but not so high itd miss the action &#55357;&#56859;

I'd be concerned about anything that could be easily mounted getting nicked. You also may need permission from the council and would need to scrub number plates if you publish results. That aside, go for it.

Random thoughts on measuring Red Light Jumping

Comments