Slashdot Mirror
Bluetooth Used To Track Traffic Times
First time accepted submitter ChanukahZombie writes "The City of Calgary, AB has introduced a new traffic congestion/timing information platform for drivers. 'The system collects the publicly available data from Bluetooths to estimate the travel time and congestion between points along those roads and displays the information on overhead message boards to motorists.' Currently only available on the Deerfoot Trail (the city's main highway artery) but will be 'expanded in the future to include sections of Crowchild Trail and Glenmore Trail in the southwest.' As for privacy concerns, the city says it cannot connect the MAC address collected to the device owner."
As for privacy concerns the city says it cannot connect the MAC address collected to the device owner.
Until they arrest someone and subpoena any data related to that person's MAC. Then they've got a nice bit of tracking data.
TFA is low on details re: what Bluetooth devices are being monitored. I know my cellphone and laptop have Bluetooth support, but I keep that mostly turned off. Do all cars in Canada come with built-in Bluetooth tracking technology? Triangulating from actual cellphone signals appears to me to be a more fool-proof if not spook-proof technology. The limited range of BT devices do make them a better choice in terms of privacy.
There, FTFY
Bits of code, random ramblings: jakimfett.com
Of course no government or private entity would ever start tracking speeds of drivers and start sending owners of the phones speeding violations if they're deemed to be speeding, right?
Of course, if you live in Calgary and you have to drive anywhere via Deerfoot, Crowchild or Genmore anywhere near rush-hour times you're painfully aware of how congested the traffic is, no need for realtime updates when there's 40 cars of stop and go in front of you.
Because everyone knows it's impossible to spoof a mac address...
Google Maps already does this with Android phones feeding google traffic density and speed data.
Its eerily accurate.
Traveling over the holiday weekend we got into some crawling slow traffic on the freeway. Google maps traffic layer said the red zone would end ahead as soon as we passed a particular location which just happened to be near a car dealership. As soon as we drove by that dealership traffic resumed normal flow.
And they do this with zero additional infrastructure. Why is Calgary wasting tax payer money installing additional sensors, when they could buy the service from Google, or probably just use it for free?
Sig Battery depleted. Reverting to safe mode.
If they only need the MAC addresses for the time that the device is traversing the system, then there's no reason to log the data for long term. TFA doesn't say how long they keep the data. Were the journalists too stupid to ask that obvious question, or did the government say "We'll get back to you"?
I have a page with links to other examples of, and discussion about, bluetooth traffic monitoring; see http://kegel.com/bluetooth-roads/
When they ran the pilot last year (or was it two years ago) I found it useful and actually was fairly accurate. Deerfoot bottlenecks and congests at certain points every day and once you are past those it moves along OK if there aren't any accidents, blizzards, etc. If I see an estimated time that is way larger than the usual time including bottleneck congestion I'll probably try a different route. I was surprised and happy to see it up and running yesterday.
Could have saved them a bundle with my competing technology, a sign that says "5:30am to 8:30pm: congested"
There is a significant density of Android phones. Just about everywhere.
I just whistled up a map of Calgary, and turned on the Traffic layer. I can see every traffic jam in the city in real time.
If you can't see that, perhaps you need to learn how to actually use your phone.
Sig Battery depleted. Reverting to safe mode.
...but come on, does anyone here really believe that if you choose to enable your Bluetooth device that others are not free to interface with your device to the extent that they can uniquely identify it? If you don't want to be tracked, maybe you should think twice about turning on Bluetooth.
I figured your non-fix deserved my non-fix.
John
Oh, that could be fun. With a few buddies you could make their system think you were going 500 MPH or so. Do it with enough devices and you could probably get their signs to say something like: Avg Time to Vancouver - 1 hour.
W..w..W - Willy Waterloo washes Warren Wiggins who is washing Waldo Woo.
I wonder how Google is able to map the traffic density in countries where Google Map navigation is not available. For example, Google Map navigation is not available here in Malaysia, so I use a modded version of the Google Map android app that allows navigation internationally. Surprisingly, I can turn on the Traffic layer in the app and it would show the traffic density. Is Google actually getting the data from the modded apps? I would be surprised if so many people here are actually using the modded version instead of the official app that disables navigation.
Dropbox drops it like it's hot.
Personally I think this is an ingenious use of technology. You people are so paranoid about privacy. You seem to be able to find a sinister side to everything, don't you? Come on, get over it. Let's celebrate creativity instead of always raining on people's parade.
Dropbox drops it like it's hot.
It seems that every comment so far has centered on the privacy implications of the collection of MAC addresses, I'm a little more concerned with where the collection of a few more bytes of data could go. I'm a few years out of the parking industry now, but the big new technology just few years ago was the use of OCR to collect license plate numbers in real time. I recall an industry presentation hawking a handheld device that could take a picture, and do the computation with little noticeable delay. It seems to me that it's not a huge technical leap to combine these, and build a super-duper tracking system. They've already tied your license plate data to your identity, after all.
A bit off-topic, but have you heard they're going to be tracking cell-phone (and accessories) signals to monitor traffic patterns? It's amazing! Why doesn't Slashdot ever accept a story on the subject?
You can read more here:
http://slashdot.org/story/05/11/19/143247
http://slashdot.org/story/05/11/19/0745248
http://slashdot.org/story/05/11/01/159241
http://slashdot.org/story/05/10/16/076217
http://slashdot.org/story/02/12/30/1243247
http://slashdot.org/story/02/06/13/0428229
http://slashdot.org/story/06/08/10/2337259
http://slashdot.org/story/07/08/31/168228
http://slashdot.org/story/12/11/28/2318245
http://slashdot.org/story/06/11/05/2220211
http://slashdot.org/story/02/10/14/1224244
There, that's better. Hopefully, one day they'll come to their senses, and post a story or two on the subject.
Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
Yes, Google is certainly gathering analytics from navigation handsets, but that is never their ONLY source of traffic flow information.
In the US, various government agencies make near-real-time traffic information publicly available. In addition, a number of private companies aggregate that info with their own additional sources, and re-sell it to other companies who need traffic information.
Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
The basic idea is that you use a set of Bluetooth receivers interspersed along a traffic corridor and attempt to track unique MAC addresses through the corridor and thus you can come up with an average, near real time, travel time through that corridor.
Some of the more interesting parts of doing include a car full of people, each with a cell phone and a laptop and quite possibly the car's own Bluetooth system. So while it is good for averaging speed and validating other measurement methods, it is not very good for counting the number of devices moving through a corridor.
Hey KID! Yeah you, get the fuck off my lawn!
[sarcasm] Outlandish as it may seem but someone might have the technology [/sarcasm]
Yet.
Shoes for Industry. Shoes for the Dead.
Is Google actually getting the data from the modded apps?
I suspect they pull that data directly out of their asses. I just checked my city. Its 5 in the morning, street in front of my house is empty yet Google maps is showing heavy congestion :)
Who logs in to gdm? Not I, said the duck.
CALTRANS uses loop detectors in freeways and major roads to monitor congestion. They just count cars in each lane and measure how fast they're going. They've been doing that for over two decades. You can see the result at . LA used to have a dedicated cable channel with that data. No privacy-invading user-identifying technology needed.
The data is used in several ways. The most important one is that when the system detects high traffic density at slow speed at one sensor, and lower density at higher speed at the next one in the same direction, it means trouble, usually an accident. The traffic detectors report the lanes separately. If something is blocking a lane and traffic is going around it, that's detected too. Cell phone and Bluetooth monitoring won't give you that.
CALTRANS has had cameras (which you can watch on line) on high poles over freeways for decades. Some have pan, tilt, and zoom capability, so when the automated system detects trouble, someone can use a camera to look at the problem area and dispatch whatever is needed.
Another use of this data is to control the metering light system at on-ramps. Freeway throughput peaks at 35 MPH (at higher speeds, the cars have to space out more) and cars are deliberately delayed a few seconds at on-ramps when speeds drop below that level.
Both of these functions require reasonably accurate data, but there's no need to identify cars individually. This all works quite well without it. Probably better. Counting all the cars on a second by second basis is more useful for detecting problems fast than some statistical measure of some of them.
The data also goes out to web sites, apps, driving time predictors, etc. There's an free API, integration with transit data, integration with CHP incident info, a developer group, etc.
A truism of traffic management is that fast response to trouble on a freeway increases the capacity by about one lane, and it's a lot cheaper than adding a lane.
So I'm not too impressed with some small-scale trial that snoops on Bluetooth headsets.
I just checked my city. Its 5 in the morning, street in front of my house is empty yet Google maps is showing heavy congestion :)
I'm guessing that you upgraded to iOS6.
It's been done in the Netherlands on dozens of locations already. Also, "anonimized" cell tower information (GSM/3G) is being used by TomTom to do the exact same in several countries.
I was promised a flying car. Where is my flying car?
why would they need that?
you only need traffic and security footage to enable them to connect it to a certain car.
anyhow, just put the bluetooth off?
world was created 5 seconds before this post as it is.
Need a few devices that randomly report Bluetooth MAC addresses ...
It's a perfectly cromulent word.
If you were blocking sigs, you wouldn't have to read this.
Some of the highways around Philadelphia now have a similar system, but I believe they may use the EZPass toll collection RFID tags to get travel times.
The system collects the publicly available data from Bluetooths to estimate
Wouldn't that be Blueteeth?
Slashdot still doesnâ(TM)t support Unicode after it was added to the HTML standard in 1997.
Indeed, it's so accurate that state DOTs are just buying data from Inrix (the same company from which Google gets its data, I think) instead of bothering to install loop or video detection.
"[Regarding the 'cloud,'] ownership was what made America different than Russia." -- Woz
That's because there are too few samples to calculate a reasonable average, so the result is essentially a divide-by-zero error.
What Google ought to do is either not show the traffic overlay for that segment, show a gray overlay, or show a confidence measurement in a tooltip.
"[Regarding the 'cloud,'] ownership was what made America different than Russia." -- Woz
The same system is being installed around Boston, MA and other localities. I love seeing the sign and knowing how long it is going to take me to reach a certain point. It takes a lot of stress out of being stuck in traffic. Less stressed drivers means traffic loosens up and moves more freely, instead of people bunching up on each other's cars and causing a traffic jam.
Additionally, the data will be publicly available, so mapping applications on GPS devices and smart phones can show traffic congestion in realtime, giving people the opportunity to plan alternate routes more easily to avoid congestion, and thereby reducing congestion.
Whew! This water sure is cold!
I love the Traffic layer. It works so good. It feels like the data is never really more than 5 minutes old, and when I'm on Highway 2 going home from Nisku to Edmonton every shitty day, I can see exactly where that day's traffic jam starts and ends, and it is like usually bang on. I use it to avoid messes all the time. So good.
How would they differentiate between people walking along the street with their phones in their pocket and BT-enabled cars driving along the same street, if all they're basing the collection on is the device MAC address?
"I love animals! Some are cute, others are tasty, what's not to like?" - Betsy Schroeder, Jeopardy contestant
Which then becomes a felony if you do so in transit for interfering with a traffic control device.
That's like an old math puzzler: if you are 30 miles from your destination, and you drive 30 MPH to get there, how fast do you have to make the return trip to average 60 MPH for the whole trip?
A lot of people will quickly say 90 MPH.
John
We must test this. For Science!
Bits of code, random ramblings: jakimfett.com
That's because there are too few samples to calculate a reasonable average, so the result is essentially a divide-by-zero error.
What Google ought to do is either not show the traffic overlay for that segment, show a gray overlay, or show a confidence measurement in a tooltip.
Its a capital of west European country. Something tells me this is not a lack of data but a crappy algo.
Who logs in to gdm? Not I, said the duck.
The city of Minneapolis also estimates travel time and displays the information via over-road screens. I'm not sure how they calculate the trip time (I don't believe it's via bluetooth), but it's a waste of effort, time, and thus money
Here's why it sucks: The display boards are infrequent - there's like 1 board along my 25 minute drive, and I drive via heavily traveled highways and freeway. Because they are so infrequent, I can only obtain trip length information for half of my trip.
Here's what they should have done: Hired a developer to create Android and Apple apps which pull trip information from Google Maps for each individual's specific route. Google Maps already adjusts the estimation for trip time when additional traffic is on the road. Google has already done the calculations - no need to reinvent the wheel.
Further more, the trip time is virtually worthless by itself. It only serves to give you a realistic expectation. Ideally, we should suggest alternate routes, for example, leaving the freeway early and driving on a less traveled frontage road. If they took off optimized the number of travelers on the freeway, it could speed up the average speed of those on (and off) the freeway.
No trees were killed to send this message, but a great number of electrons were terribly inconvenienced.
Similiar systems have been in used at airports for years, to measure waiting times at security checks etc. The ones that measure road congestion can be bought off the shelf, or can be rented (when a construction company works on a stretch of highway for instance). This is really old technology, and I wonder how it ever got up to Slashdot's front page.
Exactly his point. Swoosh!
Sig Battery depleted. Reverting to safe mode.
No, it's lack of data. Didn't you read the post I replied to?
These sorts of things collect data in 5-minute increments (maximum), so if there aren't any cars between 4:55 and 5:00 AM, which is entirely plausible even in the busiest city, it isn't going to give reasonable results. If a couple of cars then go by around 5:01, the algorithm will update and start giving good results again at 5:05.
(Un-disclaimer: Why yes, I am a traffic engineer working in ITS [Intelligent Transportation Systems].)
"[Regarding the 'cloud,'] ownership was what made America different than Russia." -- Woz
there was this case, when aol released search data together with a pseudonymous user-id. Many people got uncloaked, because first they searched for their own name or some other identifiying information, then for something like sex toys.
So the bluetooth-data isn't anonymous, too. Now they have the MAC and your traffic pattern, and when you have your device with you when you're alone somewhere i.e. in a police office, they may collect your MAC and connect it to your full name. When they later get access to the "anonymous" traffic database ... it isn't very anonymous anymore.
No, it's lack of data. Didn't you read the post I replied to?
These sorts of things collect data in 5-minute increments (maximum), so if there aren't any cars between 4:55 and 5:00 AM, which is entirely plausible even in the busiest city, it isn't going to give reasonable results. If a couple of cars then go by around 5:01, the algorithm will update and start giving good results again at 5:05.
(Un-disclaimer: Why yes, I am a traffic engineer working in ITS [Intelligent Transportation Systems].)
5 in the morning = no cars at all, maybe a lonely garbage/cleaning truck.
Capital of west European country.
Biggest/longest street in the city, going from suburbs right through downtown.
>it isn't going to give reasonable results
the F does that mean? lack of input means EMPTY, not heavily congested.
Btw checked again (almost midnight right now) and almost 1/3 of this street is heavily congested lol. Maybe googles definition of traffic is "some cars some of the time". It might also count people sleeping in their flats located along the street as sitting in traffic ...
Who logs in to gdm? Not I, said the duck.
Now that you mention it, what it could be doing is counting that garbage truck (stopping at each house) as stop-and-go traffic. If there are no other speed-limit cars to balance it out, it thinks the whole street is congested.
You're right, and that's exactly the problem that was trying to explain.
For what it's worth, other systems -- such as the one I use at work -- do report confidence levels, and do report nothing (not "empty," as it's impossible to distinguish between a road with no cars and a failed detector, but rather no result at all) when there's no traffic.
"[Regarding the 'cloud,'] ownership was what made America different than Russia." -- Woz