Again, we send our undercover cyclist and pedestrians into the city streets. Unlike in the Panorama documentary, we aren't going to make them do some in-mirror filming where they cry about the hard time they got -that's because we don't care about their feelings. Instead we just watch with bemusement.
Here, then, is an evening crossing of Montpelier. The bike is apparently equipped with LED head lamps, though this doesn't show up well in the video. Fairfield Road, a zigzag on Richmond Road then a descent on York Road -with oncoming traffic- followed by a drop down Brook Hill and along Upper Cheltenham Place to join York Road again.
What does show up is that there isn't much through traffic, and what there is gives way to the bicycle. It's hard to say that Bristol persecutes cyclists when every one of the approaching vehicles comes to a halt. The photographer claims that this is due to the fact that "My LED helmet lights cause physical pain to drivers I look at", which is something we aren't sure about, though it is certainly possible that they may be stopping because they can't actually tell that it is an oncoming bicycle. We would have to do some A/B testing with different lighting options to reach a conclusion. For now we shall just assume that in this city everyone is gracious to bicycles, even though they are clearly inappropriate for these narrow city streets.
Note how at the end of the ride, the cyclist has the rudeness to tell off a car for driving round without lights. What's sauce for the goose is sauce for the gander we say, and if bicycles are going to nip round our streets without lights, not stopping at traffic lights or zebra crossings, then we car drivers should have the same privileges -and that includes the option of driving without lights. Its more environmentally friendly, you know.
However, that's not what we really want to cover. What we do want to raise is that Montpelier is simply too narrow for cars to get past bicycles, and that Picton Street-York Road-Fairfield Road route is not some residential road, it is the primary rat-run of Montpelier, cutting out two sets of utterly needless traffic lights and two mini roundabouts. If you want to get from Stokes Croft to Ashley Down hill, you are in a hurry and you know the back roads, this is the secret route to take. Yet all it takes is one single bicycle and there's suddenly a tail back of traffic, and all that time advantage is lost. The cars parked on either side of the road are as far over as they can be, and now with the police sending them threatening letters, it's going to be hard for them to provide any passthrough space for bicycles.
We should ban bicycles from those roads! This is a key through route in the city for well-informed locals, not some back route for bicycles. They should be forced to use Cobourg Road and Upper Cheltenham Place or Shaftesbury Avenue, not this, the main road through Montpelier. Especially during morning and evening rush hours. Banning bikes here would make it much easier to get a car over to St Werburgh's and then onto the Mina Road. Here's our proposal -red is the red-marked car route, green for the tax dodgers:
View Priority Routes Through Montpelier in a larger map
Those drivers trying to use this route are important revenue earners for the city -pootling cyclists aren't. And independent of whether or not this area gets 20 mph zones, oncoming bicycles delay cars here, and that costs all those people money. Ban them from the main streets in Montpelier -give them the back roads- and all will be well.
Showing posts with label camera. Show all posts
Showing posts with label camera. Show all posts
Thursday, 29 October 2009
Thursday, 12 February 2009
Not fearing the Watchman
It turns out The Watchman, our new ANPR-camera system on Whiteladies Road has a blind spot.
If you park in the bus lane right underneath it, the camera doesn't see it.
If you park in the bus lane right underneath it, the camera doesn't see it.
Tuesday, 10 February 2009
Speed signs
Cotham Councillor Neil Harrison has apparently reported that the Cotham-Road speed warning sign was not working. One question: how exactly do you note that it is not working? Very few experiments spring to mind
1. Drive down the road at 35 mph or higher and see if you get told off by the sign
2. Put a speed camera up and see if any vehicles are speeding but while the sign stays unlit.
If the camera actually counted vehicles/day and illuminations/day, then we'd have an interesting data source for traffic volumes and the percentage speeding- and you'd have an easy way to detect failure. But sadly, this is not a feature of these signs, which is a pity. It would then be of datamining value, rather than just decoration.
Now that the sign is apparently fixed, we need a test for that too. Again, the test is simple: drive down the road fast. In a five minute period on Saturday afternoon, 30% of the passing cars appeared to be participating in this test, which shows how much they care about signs working.
Here, for example, is the Audi A055GEY, showing that whatever speed it was doing, it was enough for the light to come on
A lull in traffic provides enough time for some pedestrians to cross. This crossing here is the main crossing point to any schoolkids going to the nursery school on the right-hand-side of the road (mostly from cars temporarily parked in the buses-only parking on the LHS), and for secondary school-age kids going to Cotham Grammar across the road.
With the school rebuilding going on, there is likely to be more pedestrian traffic here. Hopefully the 30 mph sign and a flashing 20mph sign on weekday mornings is enough to ensure safe crossings. That is if the sign works.
Again, a car volunteers to check the sign lights. It's not quite so obvious from the camera angle, but whatever speed it is calibrated too, it has been triggered by a Peugeot.
A period with the sign off would actually be a good basis for a "do these signs actually work" experiment. Turn it off for a month, measure traffic volume and average/peak speeds. Turn it on again, measure the same numbers. See if it has any effect whatsoever.
Correction: The first post called the councillor Neil Hamilton, who was in fact the Conservative-cash-in-an-envelope MP for Harrods; Neil Harrison is the LibDem councillor for Cotham. He says he determined the sign wasnt working it by watching it for 15 minutes as cars went past, and noticing that it never went on. Yes, that would work.
1. Drive down the road at 35 mph or higher and see if you get told off by the sign
2. Put a speed camera up and see if any vehicles are speeding but while the sign stays unlit.
If the camera actually counted vehicles/day and illuminations/day, then we'd have an interesting data source for traffic volumes and the percentage speeding- and you'd have an easy way to detect failure. But sadly, this is not a feature of these signs, which is a pity. It would then be of datamining value, rather than just decoration.
Now that the sign is apparently fixed, we need a test for that too. Again, the test is simple: drive down the road fast. In a five minute period on Saturday afternoon, 30% of the passing cars appeared to be participating in this test, which shows how much they care about signs working.
Here, for example, is the Audi A055GEY, showing that whatever speed it was doing, it was enough for the light to come on
A lull in traffic provides enough time for some pedestrians to cross. This crossing here is the main crossing point to any schoolkids going to the nursery school on the right-hand-side of the road (mostly from cars temporarily parked in the buses-only parking on the LHS), and for secondary school-age kids going to Cotham Grammar across the road.
With the school rebuilding going on, there is likely to be more pedestrian traffic here. Hopefully the 30 mph sign and a flashing 20mph sign on weekday mornings is enough to ensure safe crossings. That is if the sign works.
Again, a car volunteers to check the sign lights. It's not quite so obvious from the camera angle, but whatever speed it is calibrated too, it has been triggered by a Peugeot.
A period with the sign off would actually be a good basis for a "do these signs actually work" experiment. Turn it off for a month, measure traffic volume and average/peak speeds. Turn it on again, measure the same numbers. See if it has any effect whatsoever.
Correction: The first post called the councillor Neil Hamilton, who was in fact the Conservative-cash-in-an-envelope MP for Harrods; Neil Harrison is the LibDem councillor for Cotham. He says he determined the sign wasnt working it by watching it for 15 minutes as cars went past, and noticing that it never went on. Yes, that would work.
Tuesday, 6 January 2009
The Watchman has arrived
When discussing what looked like a bus lane camera, James Barlow, author of Bristol's second best web site, noted that it was a watchman. More research was called for.
The watchman series of cameras include ANPR, 3G data modem, high-resolution video camera and a local hard disk drive capable of storing number plates (2 million, they say), and video shorts. If you are speeding, they actually get a video. They can also do bus-lane violation checking, which may worry the Peugeot 307 seen doing exactly that. To verify that a car shouldn't be in there it is not enough to say "is it a car"; you need to check it isn't a taxi or minicab. The Association of British Drivers have a list of proposed allowed vehicles, including lifeboat crews. Before sending a ticket to the '307 driver, the council will have to verify that the drivers/passengers were not on a lifeboat callout, on their way to Bristol docks.
The Watchman is not designed for simple speed-traps; it is a monitoring tool. It looks like it may be able to measure traffic numbers and all vehicle speeds, which could be used for datamining. As it is not currently part of the national ANPR network used to monitor and log all motorway traffic "for homeland security", this may be data we could get hold of.
The camera comes with software that sends polite notes warning drivers of their behaviour; warning notes. Their purpose is to remind without penalising. If you get a ticket, you have to pay, the insurance company ups their charges, and you may get disqualified. If you get a polite note, you can ignore it, unless eventually you start to get tickets.
Will it work? The device itself can find out. Start by monitoring peak traffic speeds before the notes begin being sent out, and their registration numbers. Then start sending notes out, and see if peak speeding speeds drop -or if the previously noted vehicles get seen repeatedly. Again, this could be a simple datamining exercise.
The sales brochure implies the cameras can be set to look for specific vehicles; download a number in and the camera will send an alert out when a specific car is seen. With a hard disk, you've got room to store a lot of numbers to search through.
Security wise, the over-sturdy mounting point is probably designed to resist a truck driving into them and getting away. You can still drive a truck into them, but it will leave noticeable damage on your vehicle. The manufacturers are proud of their security measures, though they also give away the windows software needed to connect to the cameras. The attack points for this devices are no longer the yellow box, they are the network. The camera will have a mobile phone number and a password to guess; the computers used to access it will have to stay up to date with Windows monthly OS patches or they become a target in their own right. That's the problem with networked applications: you think your device has just discovered the Internet, but in reality, the Internet has just discovered your device.
The watchman series of cameras include ANPR, 3G data modem, high-resolution video camera and a local hard disk drive capable of storing number plates (2 million, they say), and video shorts. If you are speeding, they actually get a video. They can also do bus-lane violation checking, which may worry the Peugeot 307 seen doing exactly that. To verify that a car shouldn't be in there it is not enough to say "is it a car"; you need to check it isn't a taxi or minicab. The Association of British Drivers have a list of proposed allowed vehicles, including lifeboat crews. Before sending a ticket to the '307 driver, the council will have to verify that the drivers/passengers were not on a lifeboat callout, on their way to Bristol docks.
The Watchman is not designed for simple speed-traps; it is a monitoring tool. It looks like it may be able to measure traffic numbers and all vehicle speeds, which could be used for datamining. As it is not currently part of the national ANPR network used to monitor and log all motorway traffic "for homeland security", this may be data we could get hold of.
The camera comes with software that sends polite notes warning drivers of their behaviour; warning notes. Their purpose is to remind without penalising. If you get a ticket, you have to pay, the insurance company ups their charges, and you may get disqualified. If you get a polite note, you can ignore it, unless eventually you start to get tickets.
Will it work? The device itself can find out. Start by monitoring peak traffic speeds before the notes begin being sent out, and their registration numbers. Then start sending notes out, and see if peak speeding speeds drop -or if the previously noted vehicles get seen repeatedly. Again, this could be a simple datamining exercise.
The sales brochure implies the cameras can be set to look for specific vehicles; download a number in and the camera will send an alert out when a specific car is seen. With a hard disk, you've got room to store a lot of numbers to search through.
Security wise, the over-sturdy mounting point is probably designed to resist a truck driving into them and getting away. You can still drive a truck into them, but it will leave noticeable damage on your vehicle. The manufacturers are proud of their security measures, though they also give away the windows software needed to connect to the cameras. The attack points for this devices are no longer the yellow box, they are the network. The camera will have a mobile phone number and a password to guess; the computers used to access it will have to stay up to date with Windows monthly OS patches or they become a target in their own right. That's the problem with networked applications: you think your device has just discovered the Internet, but in reality, the Internet has just discovered your device.
Tuesday, 9 December 2008
Bus Lane Camera
A new camera has popped on Whiteladies Road. Discreetly grey, with a bit of yellow on the back. It appears set up to catch cars going down the bus lane here. Presumably this will stop FirstBus buses being delayed on their morning journey into town, and reduce their need to run over bicycles.
Assuming, of course, that the problem is moving traffic in the way. Because a single day's study of the lane implies that camera or no, most car drivers aren't going to go down this lane. Why not? Parked cars in the way. The lane is bus-bike-taxi-motorbike only in the mornings, parking rest of day, so the lane is marked out with parking spaces. Which acts a cue: you can park here.
Here, at 08:50 on a weekday morning we see a volvo
A *ticketed* Ford Escort
And this VW Polo, which has pulled over to take a phone call. For this we must praise them. It is clear from these photos that there are no buses in sight -hence none being held up- and bikes can slide past without having to change into the other lane.
Why then, the camera? What it could be used for is to detect parked cars. If it counts vehicles going past, and that number is zero, there are cars parked down the lane so the attendance of Bristol Parking Services is required. Except, how can you distinguish "zero vehicles cars in the way" from "zero vehicles firstbus 'light' schedule". You can't, can you?
Assuming, of course, that the problem is moving traffic in the way. Because a single day's study of the lane implies that camera or no, most car drivers aren't going to go down this lane. Why not? Parked cars in the way. The lane is bus-bike-taxi-motorbike only in the mornings, parking rest of day, so the lane is marked out with parking spaces. Which acts a cue: you can park here.
Here, at 08:50 on a weekday morning we see a volvo
A *ticketed* Ford Escort
And this VW Polo, which has pulled over to take a phone call. For this we must praise them. It is clear from these photos that there are no buses in sight -hence none being held up- and bikes can slide past without having to change into the other lane.
Why then, the camera? What it could be used for is to detect parked cars. If it counts vehicles going past, and that number is zero, there are cars parked down the lane so the attendance of Bristol Parking Services is required. Except, how can you distinguish "zero vehicles cars in the way" from "zero vehicles firstbus 'light' schedule". You can't, can you?
Subscribe to:
Posts (Atom)
Posts
