How many of you carry bluetooth enabled/activated phones or gadgets?
This is a serious question.
There are devices on the market that are used in many large cities to monitor highway congestion by monitoring the progression of bluetooth devices along a given route.
Essentially, the “bluetooth reader” is a small box that picks up an EZ pass device, cell phone or other blue tooth enabled device as it passes.
A mile or several miles down the same corridor a second box will also grab bluetooth signals as the vehicle passes. There may be third, fourth and additional readers further along the highway or the corridor.
By matching the bluetooth tags as they move along the corridor, the “device” (as opposed to the DOT or other implementation agent) can monitor traffic speed, identify congestion, etc.
This (or something similar) is used in major cities all over the US as the basis for morning traffic reports, etc.
When discussing the program with a colleague recently, the question of “false readings” came up. In part, the system is designed to kick out anyone who takes too long to complete the link between two readers. So, if you stop for gas, or to drop your child at school, your bluetooth readings won’t be read as a “match” because you exceeded the time threshold.
But, what about slow vehicles? Would a blue tooth reader pick up a device from a cyclist? What would that device be? A cell phone most likely, but are there other devices that you might ride with that is blue tooth enabled? And would that blue tooth device be enabled as you were riding?
What do you ride with that would be blue tooth enabled, and blue tooth activated, when you ride? Your input could be very helpful.
If you apply just “taking too long to transfer” you will miss a lot of heavy traffic. You need count how many registered devices take too long and if, for example, one or two out 100 then just drop them. If this is for 50 then it’s a heave traffic in some direction. Numbers are out of blue and simple threshold is probably too simple but you should get an idea.
Android smartphone. And I’ve been geek-aching to have a bluetooth device/application that was effect-justified.
There are the helmet g-force sensors that notify 911 when you’ve been knocked down. If it senses G’s followed by no/low motion, it starts calling your cellphone – “Hey, I’m going to call 911 in x minutes…” and then within a time parameter it auto-dials 911 and sends out whatever text messages to whoever you pre-designate. Hello Mom, I probably just dropped my helmet and my cellphone BUT this is that message we talked about..
I currently have a two-user bluetooth comm system for bicyclists that I’m going to evaluate. That connects both riders phones, their music players, and an intercom channel and provides a priority cascade between the channels.
I aspire to a QuantifiedSelf lifestyle and I particularly need a BlueTooth (or other technology) continuous glucose monitor, and there’s none on the open market yet. Man I want one of those.
Certainly, you could see that some cellphone counters placed over the trails at major crossings could automate the semiannual traffic counts and provide much better data. I mean, Homeland Security is doing it. (they call it a “Stingray” device)
Sorry- too much? V.
I always ride with my iphone. Bluetooth is on 95% of the time.
Man, I’m going to start turning off my phone unless I’m actually using or waiting for a call.
i dont use the bluetooth on my phone, but i do use the gps.
its great fro tracking rides then uploading onto the web for analysis & stuff. (too bad my battery is garbage and it usually only lasts from oakland over to southside.)
@Edmonds – suddenly opposed to crowdsourcing, are you? I am surprised.
@mikhail – your concerns are valid. But, in an over stressed system of the sort you describe, the data points would drop to zero, as all cars/devices would exceed the permitted time window. Common sense would then dictate reliance on secondary sources for data, such as traffic cameras. And, the location and circumstances would determine the window. For example, there should be few circumstances in which it will take you more than XYZ minutes to get from Camp Horne Road to Bellevue. If you need to leave inbound I-279 long enough to get gas in the morning, and bluetooth readers are in place near each exit, for discussion purposes, then the amount of time that it takes you to exit the highway, fill your tank, asnd then re-enter the highway and progress to the next bluetooth reader should a) kick you out of the pool for that road segment due to your much lower than average speed or b) make you a low speed data outlier. So, it would affect overall data only if LOTS of people stopped to get gas, thereby artificially lowering the average, or b) there was an unacceptably low number of bluetooth devices to be read, in which case the average “value” attributed to this outlying data is magnified.
My deepest apologies to all the statisticians out there. I realize that my “explanation” leaves much to be desired from a statistical standpoint. Please be kind. My last statistics course was probably 25 years ago or more! I think I remember the basics, tho! Or I hope I do.
I proudly do not own a cell phone or any other type of mobile device. I carry a pencil and paper if I really want to remember something (license plate, grocery item, etc.). If I want to talk to someone I wait until I see them or until I decide to call from my land line or use email. If someone wants to reach me they can do the same.
I usually carry a phone, but I infrequently use it. I keep it turned off most of the time. I always keep blue-tooth off unless actively using it. Wifi is also turned off.
I am totally in favor of crowdsourcing everyone else. Ha! ;)
I should point out that various companies have more or less figured out the details of how to do this routinely. Waze does it (and gives specifics). Google does it (just tick Traffic). Garmin does it (and broadcasts the results to your device).
As for all such things, you collect data and you build statistical models. The models allow you to interpret the (noisy) raw data and produce quite accurate information.
Quite right. You don’t need all the bikes to have a cellphone to do a traffic count.
You do a series of parallel observations, collecting both digital and visual counts, and determine the ratio of cellphone-active-bikes.
Then you just conduct electronic surveillance OOPS automated counting, and apply the correction factor, and you do a visual audit periodically to make sure your multiple is reasonable.
The really cool thing to my mind is if you were stealth-collected cellphone data (like DHS does), you wouldn’t just have numbers at intersections; you could build pattern numbers and know how many people at 1500 ride Millvale to Southside, etc.
I’m wondering if you don’t want to visit Google-Pittsburgh and talk to some of their folks about a 20% project to help analyze local bike flows. You don’t want any real user data, you just want aliased/anonymized tracked data. Because Google already has all the data you want.
You must be logged in to reply to this topic.