Cycling in Sydney Australia
On the weekend I decided to go and ride the full length of the Bourke St cycleway as it's not really on my regular routes. I'd ridden short sections of it before, but never the newer bit north of Taylor Sq, since it doesn't really go anywhere (unless you're after a Pie'n'peas).
There is definitely a huge problem with the traffic signals. Every single bike signal bar one, failed to detect my bike for the full length of the route out and back. This is completely dysfunctional and is in fact this is probably its biggest achilles heel.
I stopped over the sensors. Sat through multiple phases as I wasn't in a hurry. Nothing.
As it was the weekend, I was on my "nice" bike, which has a carbon frame, but the wheels, cranks and pedals are metal. This is a pretty common setup, so if the signalling system cant handle that, it's a failure.
The other problem is inconsistent treatment of right of way for traffic crossing the cycleway. At some points vehicles coming from the road must give way to the bikeway, and others not. There needs to be a single standardised treatment of this scenario. Inconsistency is hazardous.
Similarly, it is unclear whether the crossing at the Eastern Distributor entrance applies to bikes?
Highlight was seeing an older gent accidentally ride the steps immediately north of Oxford St. He thought he was following the cycleway, but it deviates off to the left. He went straight! He stayed on though, and despite looking a little rattled, said he was ok. Awesome.
There's no question that it has improved the streetscape, and had a calming effect on the traffic.
It's great for a quick trip to the bakery etc, especially if you don't have to cross to many intersections.
The surface was clean and free of glass.
I ride this section all the time down as far as William Street, and sometimes down to the bay itself on weekends.
Usually don't have any problems with the sensors along that part. The only sensor I can't get to work is the Liverpool St northbound (downhill), but the general traffic lane sensor picks me up there so I use that.
Apart from the sensor above they all picked up my 1998 Mongoose VRS3 Mt bike, 2006 Bauer Pyrenees semi-carbon roadie and my new 2012 Giant Defy 1 commuter. I weigh about 75 and the bikes usually aroung the 11-15kg mark. If they are not picking up full carbon bikes then that is a massive fail. In any event they are so sensitive that most people don't trigger them regardless of what they ride, so they are clearly not user friendly. But then a lot of people I see stop well in front of them in the intersection itself to wait for the green so they will never work in that situation.
I think the ED crossing is such that technically you need to get off and walk across it. Meh to that.
I read in the Bourke Rd (Alexandria) independent report yesterday that bikes along there at least must give way to cars at side streets, but along the Surry Hills part there are give way lines facing drivers. Pick an intersection treatment at random and go with that it seems.
They are induction loops. ie weight doesn't matter - only proximity of moving metal to loop - rolling an aluminum wheel down the center front "T" of the loop will usually do it - in the cases where the RTA hasn't conveniently painted/filled/otherwise hidden the location of the loop.
As far as bourke st seeming broken, this will occur because the decision on whether or not to allow the cycle phase will occur somewhere between 8 and 25 seconds before the cycle phase occurs. In practice this causes the traffic lights to offer a longer average delay to cyclists than any other users, and will not offer the cycle phase when you expect it to - ie when the road alongside you is given the green.
This is an example I've posted before of it occuring on Kent St. note that the lights on both Kent and King still do this - ie they replace the cycle light with an allred phase that is the same length as the cycle green if the cyclist turns up just before the orange.
Some of the cross streets on bourke have much longer timings than a midnight kent run, so they will seem proportionally more broken.
Yep, this video shows one aspect of the problem well - you only get a green if you arrive a long time before a potential green. In determining the balance between turning motor traffic and straight-ahead bike traffic it's as though the traffic engineers have carefully weighed one side of the argument.
I really don't think there is any induction loop going north on Bourke at Liverpool St. Usually there is a scoring of the ground - but at those lights there are none.
They can be buried. However it wouldn't suprise me if the induction loop was missing.
Where Clarence and Kent join, there is a double intersection, which is capable of being open for cyclists coming off the SHB without being open for cyclists to cross the second part of the intersection. There is no induction loop fitted there either, you cannot trigger the lights if caught in the middle.
Herzog / JasonB, I also had a very poor success rate with the lights along Bourke St, I very carefully ensured I rolled directly over the "diamonds" or centre line of the loop but generally then rolled forward of the loop area so I had a good view of traffic to ensure it actually stopped, I was lucky to get 30% detection (many many times).
Then I saw a post by Andrew K wherein he described his technique (can't find the post but generally included staying within the loop, not rolling out of the loop area and placing at least one foot on the ground in the loop area) and his reported usual success. I also pointed out that this made no sense as they were "induction loops etc, but I tried it and now get almost 100% success. Still only 6 seconds etc and predominant priority for cars, some intersections like Albion & Bourke make no sense at all but Andrew K's technique at least maximises detection for me.
Honestly, having to perform like a trained monkey in order to get a princely 6 seconds of green light seems way to similar to keeping my eyes downcast, intoning "Yes Massa", and keeping to the back of the bus. I ignore the bicycle lights on Bourke St and make my own decisions about when it's safe to go (with the exceptions of Cleveland, William and Oxford Sts, which are too busy/dangerous for me to rely on my own judgement).
Is there merit in supergluing a small piece (eg: 50c coin size) of checkerplate to the surface above each sensor?
Basically the idea would be to make the system always register the presence of a "bike" and therefore run a green bike signal as part of every phase?
Usually won't work, since the induction loop is a part of an oscillator and it is the frequency shifting that is considered to be a vehicle.
So a constant shift in frequency won't trigger.
so that's why taxis reverse and then go forward again at traffic lights? I always thought it was urban myth...
Back when I was young and very foolish we used to take the reverse light globes out of our cars and trigger the old type red light cameras when no one was behind by slightly rolling over the white line loop, then rolling back behind the line again. FLASH.
Just beware those ever-vigilant surry hills bike cops...