Have you ever been driving down the interstate and you come up to a car driving 10 mph under the speed limit in the left lane? You try all your favorite hand signals but this person just doesn’t figure out that they are holding people up. Ah yes, the driving-cell-phone-talker. Studies are showing that people have similar reaction times to drunk drivers when talking on cell phones while driving. But what if we could harvest a social good from the behavior?
Nearly everyone has a cell phone with them all the time. If you were to go down the street and take a survey 80%* of Americans own a cell phone, and the other 20% is in diapers or was alive to see the inauguration of FDR (or both). The percentage among people who comprise rush hour traffic is even higher. Why can’t we use this vast data source to figure out what’s going on with traffic without adding additional infrastructure costs?
First a quick introduction on how cell phones work:
Your cell phone is a small, digital radio transmitter. Cell phone towers scattered throughout the city are what allow your phone to get a signal. There are a lot of these towers, as the phone needs to be within 2 miles of a tower to communicate. So when a carrier is telling you that they have more bars in more places they are bragging that they have a higher tower density (or feeding you a line of marketing bull)**. Even when you are not talking on the phone, you cell phone periodically sends a signal to the towers that can hear it to register itself with that particular tower or towers. This registration is tracked on the carrier’s network so that if you receive and incoming call, the system will know which towers to use to attempt to communicate with your phone. Without this, the incoming call would have to be sent over all towers all over the country, which obviously isn’t efficient. When you are actually on a call, your phone is in continuous communication with the tower to transmit and receive data.
Now on to how this information can be used to track traffic flow. Because the carrier’s network is receiving periodic updates from your phone, regardless of if you are in a call or not, the system can track an approximate location for your phone. At worst, it could track if you are moving from tower to tower at a regular speed, implying that you are in a moving vehicle traveling at a reasonable speed. Your position can be calculated even more accurately, however, as your phone will most likely be able to reach several towers from the same position. The relative time at which the signals are received by the towers (or the relative strength of the signal received by the towers) can be used to triangulate your position***. This is how phones without GPS built in can still triangulate their approximate position (e.g. the first gen iPhone).
The data from one phone is not going to be accurate enough to accurately predict uniform movement, but when you aggregate hundreds or thousands of phones in the area, plus use the known locations of roads as an assumed path when positioning moving phones, you should be able to reconstruct traffic flow on large roads with reasonable accuracy.
There are other ways of using information coming from the vehicles to determine traffic. One approach suggests using smart phones with GPS to track traffic, but this approach requires an opt-in for users having software on their phones to help the process. Constantly running your GPS in the phone also hurts battery life. Another approach already in use is to have in-car GPS units transmit their position data back to a central system via the cell phone network. This can work, but there needs to be enough cars using the GPS driving aids to be accurate in an area. With this approach you also have the problems where the data coming back is likely to be only fed to users of that service (e.g. TomTom users get data from other TomTom systems, but don’t share with the Magellan users).
With raw data from the cell network, you don’t run into these problems. The cell phone carriers aren’t in the business of selling driving optimization devices, so they are likely to be willing to sell data to municipalities to be made generally available. Phone battery life won’t be affected because cell phones need to talk to the towers at regular intervals anyway. Users don’t need to opt in so there is an endless stream of data readily available as soon as someone throws the switch.
There are even more advantages to this approach. For rural areas where it wouldn’t make sense to track traffic using traditional means, one could use this approach to detect traffic congestion in areas where you wouldn’t normally monitor for it. For example, during the summer construction season, such a system could detect slowdowns anywhere in the country without requiring construction data to be entered by each individual state or city.
Cow-collision traffic jams automatically detected!!
To take the idea a step in the direction of Big Brother, the data could be used for long term transportation infrastructure planning. By understanding how people are flowing from one area of a city to another at various times of the day, one could better understand how to improve roads to make transportation more efficient.
There are many ways the above technique for the real time monitoring of traffic flow could be used. The big benefit of the approach is that no additional infrastructure is required, and users are not required to have special phones or software installed on their phones. The system just takes existing data and uses it for something useful.
* I reserve the right to make up, or use unreliable sources for my statistics.
** Or they could be arguing that their wireless frequency penetrates buildings better.
*** With three towers, you can get an exact position; with two a line, etc.
