BIKEPGH MESSAGE BOARD ARCHIVE

« Back to Archive
13

speed cameras

A year ago, we were talking about speed cameras in this thread. http://localhost/mb/topic/trib-opinion-piece/ I didn't want to resurrect that thread directly, but it's a good lead-in to the topic at hand. In the year since, has anyone either spent the $2 for the iPhone app (or Android equivalent), or borrowed Bike-Pgh's portable speed monitor? I'm curious to know anyone's experience before I spend money and time to do it myself. My impetus is further thinking about my experience with the guy on Perrysville a few weeks ago. It would be really cool if I could show local law enforcement that [name your location] had N% of people exceeding the posted speed limit by 5 mph, and M% exceeding it by 10 or more mph.
stuinmccandless
2014-12-13 11:59:42
Do we have any indication that any Police department cares about speeding?
sgtjonson
2014-12-13 12:58:44
I think it is up to us to be the squeaky wheel, and to do so with data. Otherwise police only care after someone gets killed, and then only for a few weeks after the headlines fade.
stuinmccandless
2014-12-13 13:27:27
I just did this. Very quick and dirty test from my window. It clocked 5 cars averaging just over 20 miles per hour. That might be a bit low (2 - 4 mph estimate) but that's probably more of a function of the numbers that I used to set it up. I want to try this when I can be more precise about the distance setting.
sew
2014-12-17 10:59:48
stu: what app are you referring to?
paulheckbert
2014-12-17 11:10:07
FWIW: I downloaded "SpeedClock" on the iPhone.
sew
2014-12-17 11:11:31
iPhone SpeedClock, same as @sew. Review here. I haven't done this yet, but wondered how anyone else fared with it.
stuinmccandless
2014-12-17 12:49:15
I've been thinking about it and it should be possible to determine the speed of passing cars using the microphone on a cellphone, from the Doppler shift of the roadnoise as they pass. That together with your known bicycle speed should work pretty well as an estimate of how fast people are passing you, at least when they do it one at a time. That would work better on a bike than trying to point your cellphone or a radar gun at a passing car. Say, http://home.iitk.ac.in/~rhegde/papers/46.pdf. Hey, those guys are at CMU. I wonder if they have coded a mobile app...
jonawebb
2014-12-17 15:55:14
Looks like a fun thing to write a DSP paper on but probably not so practical in the real world. The first problem is accurately filtering the audio sample from a moving bicycle in a noisy environment. The second (and I think larger) problem is that the speed of the car and the bicycle is not constant. (And if the car changes gears...) The best bet (without radar or a laser speed gun) is probably still d/t. Pick two points, calculate the time between them, calculate the distance between them, and divide. A bicycle is an accurate distance measure as well. Just measure the number of wheel revolutions between the two points and multiply by the circumference.
andyc
2014-12-18 14:34:24
Quoting from the article: "DeFebo describes the program as one which works similarly to an EZ-Pass scan or red-light camera. When a car drives past the camera mounted in the work zone, it will take a snapshot of the car's registration while clocking its speed. If the car is going faster than 11 miles per hour or more faster than the listed speed limit, a ticket will be mailed to the offender's address." Senate Bill 840. It's a five-year pilot program, and only on the Turnpike, and only in construction zones, but it's something. Seems like a long time, but it's not that long. Five years ago was when we created Flock of Cycles, also when Don Parker got killed; that doesn't seem that long ago.
stuinmccandless
2015-10-08 16:41:19
Manchester UK speed cameras nabbing 73 a day, on average. [link]
stuinmccandless
2015-10-22 02:01:39