Friday, March 5, 2010

Seeing The Light - The Clear Need For Bicycle Detection By Traffic Signals


A pet peeve of mine I've mentioned at least once before on the blog is that traffic light sensitivity is not standardized, with some intersections picking up bikes, but most will not. It's always a surprise until you memorize the signalization of a particular route. The real kicker to me is there are many intersections at streets with bike lanes that were obviously not calibrated with bicycles in mind, even though there is a sign right there with a bike on it. The CVC 21450 seems to suggest that all streets which are repaved, and thus have their magnetic traffic signal detectors ripped out, should have the sensors reinstalled to detect motorcycles and bicycles. Legislation like seems more like a mumbled suggestion than a mandate, because I can recall repaving of several streets along bike routes in Santa Monica where bike detection remained an issue or got worse after repaving.

Signal detection is not something talked about as often as things like painting more bike lanes, but I think it is essential to legitimizing cycling for transportation and more importantly for safety. Lack of signal detection is most frustrating late at night, when the absence of automobile traffic means some intersections no longer change between red and green regularly, but may hold one color indefinitely until triggered. Since most cyclist fatalities are at night, and located at intersections, we should be paying close attention to this issue. The unchanging traffic light creates a situation where the infrastructure it self is telling a cyclist you do not belong, made especially confusing when sitting in a bike lane that says you belong, but will not detect you at the intersection.

At such an intersection a cyclist is presented with essentially 4 options which are all problematic and either frustrating, dangerous and or illegal by the letter of the law.

If the signal loops just won't budge for your bike, you can roll around and hop over to the pedestrian crossing button, assuming one is available. If you are positioned to go straight through the intersection or even worse already positioned to make a left turn, than you will likely need to roll through the perpendicular cross walk to get to the pedestrian signal, and hop off and futz the bike around and hop back on. Imagine how people would feel about driving if coming home late from the office meant having to stop the car at every other intersection, get out of the car, push a button, hop back in, and proceed to repeat this dance over and over.


Wait for a car to show up going the direction you need to travel in. If it's late at night on a low traffic road, the kind that are most ideal for cycling on, good luck waiting, it'll be a long night.


Go through the red light. This is for obvious reasons the riskiest of the 3 options, and technically illegal (However I would define such a light as defective, and there is some ambiguity to the legality of this). I will admit quite plainly that I run red lights, and over the past couple months when I was working overtime nearly every day, I ran a red light at least once on my way home nearly every day. This is the sort of thing that gets anonymous motorist internet commentators and radio shot hosts all riled up, to fathom such wanton and regular disregard for the law (despite their few qualms about chatting on the phone, California stops, and habitual speeding on their commute). The fact of the matter is, if I am working overtime and riding home tired, I am not going to sit and wait forever in the degrading process of hoping for a motorist to come escort me across at each intersection with a bicycle traffic detection problem. I'm also not going to get off my bike and go over and push a button. So I go against the red, looking both ways first of course.

Option 4:

Ride onto the sidewalk so as to approach every intersection within convenient reach of the pedestrian cross signal. A lot of people do this, even on streets with bike lanes, in large part
because our bike lanes don't feel especially safe. The problem with this is that cyclists riding on the sidewalk exposure them self to lots of potential blind side collisions from drivers pulling out of driveways and making turns at intersections, and thus make up a majority of total bicycle crashes. Not to mention if someone were to actually be walking on what is most likely a a space barely adequate for walking, let alone vehicle traffic, we have an obvious conflict between pedestrians and cyclists. Top this all off with the fact that in the City of Santa Monica it is illegal to ride on a bike on sidewalks.

One last option that only exists at a few intersections in Venice where it is installed, but there are such things as crossing buttons facing the street especially for cyclists to press without having to get off the bike or get onto the sidewalk. This is considerably easier than a cyclist having to use a conventional pedestrian button, but would be unnecessary if the signal detection was setup correctly. Lastly this option ignores the dilemma of a left turning cyclist, who would position them self much too far away to hit such a button, and so I see it as an inadequate solution to the problem.

Simply redoing every intersection to fix this problem right away is impractical and costly, but eventually every intersection gets redone for maintenance reasons anyways, which is why the C.V.C. specifies tuning for bicycles should be done "Upon the first placement of a traffic-actuated signal or replacement of the loop detector of a traffic-actuated signal". Detecting bikes is not some magical voodoo work of engineering, it's done successfully in many places, but sadly not most places. There needs to be a clear and consistent practice to tuning intersections for bicycle travel, and the lack of one currently is purely a matter of ignorance and indifference to our concerns, even in a City like Santa Monica that likes to pride it self as a cycling friendly place. Things like bike valet at the local farmers market are a nice gesture and a great idea to promote cycling while reducing demand for car parking, but does nothing for the safety and concerns of the everyday bicycle commuter.

There are several remedies for the situation. One is to as outlined in the CVC, setup the magnetic inductors to detect bikes with each new repaving, something Santa Monica has thus far failed to do consistently. The other thing that can be done is move toward camera pixel analysis signalization, which already exists at a few places in Santa Monica. These systems are more accurate, and easier to reprogram, however are significantly more expensive per intersection to install. However, in the long term they can pay for their extra expense by reducing future labor and maintenance issues. No longer would workers have to come out with diamond saws every time signal detection had to be redone. Replacing the need for cutting into the asphalt also exposes the intersection to less water erosion that slips through the cracks. This would reduce potholes and extend the life of the road.

As BikingInLA mentioned in the comments on my last post about traffic lights, perhaps Santa Monica's failure to comply with State law could set the stage for lawsuits or even a class action lawsuit against the city. If a cyclist rides against red at an intersection that will not detect them, and is hit by cross traffic, perhaps the city is partially liable, considering that they obviously failed to comply with the following conditions during routine road maintenance:
"Upon the first placement of a traffic-actuated signal or replacement of the loop detector of a traffic-actuated signal, the traffic-actuated signal shall, to the extent feasible and in conformance with professional traffic engineering practice, be installed and maintained so as to detect lawful bicycle or motorcycle traffic on the roadway."
It is quite feasible to place and tune actuators to detect bicycles, as there are plenty of places where it is actually done right. A standout example on the West Side is the North bound exit of the Marina bike path at Washington, where inductors are setup with bikes in mind and are painted to imply where the best placement for the bicycle wheels are for detection. Even some roads not ideally setup for detecting cyclists will in fact detect them. Where it is done completely wrong, as in will not detect a bicycle under any circumstance or placement, seems to be simply ignorance at work or lack of concern. I think the case to sue the City is especially strong where bicycle routes are clearly marked, as obviously cyclists are not only expected, but encouraged to ride there. Neglecting to maintain intersections for bicycle use along official bicycle routes would seem to me to be clear negligence on the part of the City.

There seems to be an awful shortage of concern for the safety of anyone not in a 4 wheel box by those who plan our roads. I'm not a legal expert, but I strongly feel that as a movement, cycling needs to explore how best to bring lawsuits against city and state governments. Civil rights activists have fought long and hard for years for handicap access, things like curb ramps for wheel chairs on sidewalks, and they didn't just ask nicely, they sued when necessary to get the job done. Our very own Caltrans just recently settled a lawsuit to the tune of 1.1 billion dollars to be used to repair their years of negligence in maintaining sidewalks and providing curb ramps. Asking nicely for that which should be your right only goes so far, sometimes you need to wield a stick. Perhaps until we see a critical mass of lawsuits, politicians will be content to mostly ignore us, or half ass try to appease us. Money talks, liability lawsuits cost money.


In the meantime, there are some tricks to improve your chances of detection while we wait for the state and local governments to put 2 and 2 together. Cyclist View has a great breakdown of the topic of signal detection, the basic science of how inductor strips work, and the ideal placement for a cyclist to trigger a light with different inductor patterns. The best tip I have read is that the key is to make 2 contact points with the loop for best detection. Ideally get your wheels lined up so the center of each wheel is lined right over the lines of a loop. Depending on the intersection this may be easy or difficult, and while perfect placement helps, some intersections will not detect you, no matter what you do, which is of course a huge flaw in our system, and a failure of our public works. Even motorcycles, much bigger hunks of metal than we have, have been known to be unable to trigger some lights. Sometimes dancing around a bit with your bike and rolling it back and forth over the loop will get it to work, kind of like blowing on old video games cartridges.

Extra Links & Sources:
Cyclist View
Bicycle Detection and Signalizatio, By Alyssa Sherman - Tech Transfer Program
Detection of Bicycles by Quadrupole Loops at Demand-Actuated Traffic Signals, Steven G. Goodridge, Ph.D.
Bicycle Detection Program, Santa Cruz, California
Traffic Signal Actuators: Am I Paranoid?


John Romeo Alpha said...

Thank you for this information. Those diagrams were very helpful because they clearly explain something I have been wondering about for a long time.

Evan said...

Thanks for the post. I had a really aggravating example of this last week, when I went for a night ride and was heading south on Yale in Santa Monica. The light at Wilshire was red, and there were no cars waiting to cross Wilshire at Yale. I moved around over the sensor, waited some more, then I saw the pedestrian signal on Wilshire change to the red countdown. "Finally," I thought, I'm going to get the green. But the pedestrian signal counted down to 1, then cycled back to the Walk sign. After that I gave up and rode onto the sidewalk to press the pedestrian button, then had to work my way back to getting ready to ride again. Thankfully at Santa Monica Blvd., a car was waiting on Yale to cross.

justin said...

Awesome, thanx Gary. There are even a few intersections in Santa Monica/Venice that won't pick up my Mazda Miata. It's happened before that I've been waiting, watching the red hand countdown for the cross traffic and been rewarded with the hilarious switchback-to-the-walk-signal that Evan mentioned. Tough to give up and go through the red in my 2500lb metal box. Road crews (in the city of LA but SM especially) don't seem to really have it together when it comes to surfacing, mixing, engineering, smoothing or repairing any kind of road surface. I wish I could expect more attention to be paid to sensors but I'm not holding my breath until they at least learn how to build a basic road.