MIT and Princeton researchers leverage smartphone cameras for collaborative traffic signal schedule advisory system; tests show 20% cut in fuel consumption
25 August 2011
|SignalGuru service architecture. Click to enlarge.|
In July, at the Association for Computing Machinery’s MobiSys conference, researchers from MIT and Princeton University took the best-paper award for a system that uses a network of smartphones mounted on car dashboards to collect information about traffic signals and to then tell drivers when slowing down could help them avoid waiting at lights. By reducing the need to idle and accelerate from a standstill, the system saves fuel: in tests conducted in Cambridge, Mass., it helped drivers cut fuel consumption by 20%.
The new system, dubbed SignalGuru, relies on images captured by the phones’ cameras. SignalGuru is a software service that leverages opportunistic sensing on mobile phones to detect the current color of traffic signals, to share with nearby mobile phones collectively to derive traffic signal history, and to predict the future status and timing of traffic signals.
Traffic signals are widespread in developed countries as they allow competing flows of traffic to safely cross busy intersections. Traffic signals, however, do take their toll. The stop-and-go movement pattern that they impose, increases fuel consumption by 17%, CO2 emissions by 15%, causes congestion, and leads to increased driver frustration. Drivers can be assisted with a Green Light Optimal Speed Advisory (GLOSA) system. A GLOSA system advises drivers on the optimal speed they should maintain when heading towards a signalized intersection. Should drivers maintain this speed, then the traffic signal will be green when they reach the intersection, allowing the driver to cruise through.—Koukoumidis et al.
Worldwide, only a of handful GLOSA systems have been deployed; their costly and often impractical deployment and maintenance, however, has hindered their widespread usage, the authors note. Recognizing the potential benefit of GLOSA, US and European transportation agencies have advocated for the integration of short range (DSRC) antennas into traffic signals as part of their long term vision for intelligent transportation systems. (Earlier post.) DSRC-enabled traffic signals would be able to broadcast their schedule to DSRC-enabled vehicles that are in range.
This approach also necessitates the cost of equipping traffic signals and vehicles with the necessary specialized computational and wireless communications infrastructure.
In this paper, we take an infrastructure-less approach to accessing traffic signal schedules. We propose, implement and evaluate SignalGuru, a software service that runs solely on mobile phones, predicting the traffic signal schedule without any direct communications from the traffic signals. Our mobile phones are mounted on the vehicle’s windshield, and use on-phone cameras to detect and determine the current status of traffic signals. Multiple phones in the vicinity use opportunistic ad-hoc communications to collaboratively learn the timing patterns of traffic signals and predict their schedule.—Koukoumidis et al.
There are drawbacks to such an infrastructure-less approach, the authors note in their paper:
Lack of loop detector information. SignalGuru works without access to such information, and must perform predictions solely based on the information that can be measured by available mobile phone sensors.
Commodity cameras. The quality of smartphones’ cameras is significantly lower than that of high end specialized cameras used in computer vision and autonomous navigation.
Limited processing power. Processing video frames to detect traffic signals and their status (red, yellow, green) takes significant computational resources. A traffic signal detection algorithm that runs on resource-constrained smartphones must be lightweight so that video frames can still be processed at high frequencies. The higher the processing frequency the more accurately SignalGuru can measure the duration of traffic signal phases and the time of their status transitions.
Uncontrolled environment composition and false detections. Windshield-mounted smartphones capture the real world while moving. As a result, there is no control over the composition of the content captured by their video cameras. Results from one of the deployments suggested that the camera-based traffic signal detection algorithm can confuse various objects for traffic signals and falsely detect traffic signal colors. A misdetection rate of 4.5% can corrupt up to 100% of traffic signal predictions.
Variable ambient light conditions. Still image and video capture are significantly affected by the amount of ambient light that depends on both the time of the day and the prevailing weather conditions.
Need for collaboration. The traffic signal information that an individual mobile device senses is limited to its camera’s view angle. A device may not be able to see a far-away traffic signal, or may not be within view of the traffic signal for a long enough stretch of time. Collaboration is needed between vehicles in the vicinity (even those on intersecting roads) so that devices have enough information to be able to predict the schedule of traffic signals. Koukoumidis et al. focused on a completely infrastructure-less solution that relies solely upon opportunistic communication (ad-hoc 802.11g) among the windshield-mounted devices.
The phone cameras capture video frames, and detect the color of the traffic signal using SignalGuru’s detection module. Information from multiple frames is then used to filter away erroneous traffic signal transitions (transition filtering module). Nodes running the SignalGuru service broadcast and merge their traffic signal transitions with others in communications range (collaboration module). Finally, a merged transitions database is used to predict the future schedule of the traffic signals ahead (prediction module).
The prediction of the future schedule of traffic signals is based on information about past timestamped R→G transitions i.e., information about when the traffic signals transitioned from red to green in the current or previous cycles. The prediction is based on R→G transitions, as opposed to G→Y (green to yellow) transitions, because vehicle-mounted smartphones can witness and detect R→G transitions much more frequently; when the traffic signal is red, vehicles have to stop and wait till the signal turns green. As a result, it is quite likely that a vehicle will be at the intersection at the moment that the R→G transition happens and thus detect it.
In addition to testing SignalGuru in Cambridge, where traffic lights are on fixed schedules, the researchers also tested it in Singapore, where the duration of lights varies continuously according to fluctuations in traffic flow. In Cambridge, the system was able to predict when lights would change with an error of only two-thirds of a second. In suburban Singapore, the error increased to slightly more than a second, and at one particular light in densely populated central Singapore, it went up to more than two seconds.
The computing infrastructure that underlies the system could be adapted to a wide range of applications, the researchers suggest, such as capturing information about prices at different gas stations, about the locations and rates of progress of city buses, or about the availability of parking spaces in urban areas, all of which could be useful to commuters.
Our proposed schemes improve traffic signal detection, filter noisy traffic signal data, and predict traffic signal schedule. Our results, from two real world deployments in Cambridge (MA, USA) and Singapore, show that SignalGuru can effectively predict the schedule for not only pre-timed but also state of the art traffic-adaptive traffic signals. Furthermore, fuel efficiency measurements, on an actual city vehicle, highlight the significant fuel savings (20.3%) that our SignalGuru-based GLOSA application can offer. Given the importance of traffic signals, we hope that this work will motivate further research in their detection, prediction and related applications.—Koukoumidis et al.
SignalGuru is a great example of how mobile phones can be used to offer new transportation services, and in particular services that had traditionally been thought to require vehicle-to-vehicle communication systems. There is a much more infrastructure-oriented approach where transmitters are built into traffic lights and receivers are built into cars, so there’s a much higher technology investment needed.—Marco Gruteser, an associate professor of electrical and computer engineering in the Wireless information Network Laboratory at Rutgers University
Emmanouil Koukoumidis, Li-Shiuan Peh, Margaret Martonosi (2011) SignalGuru: Leveraging Mobile Phones for Collaborative Traffic Signal Schedule Advisory (MobiSys’11, June 28–July 1, 2011, Bethesda, Maryland)
a: Led traffic lights could broadcast a countdown signal to tell cars when they were about to turn red / green.
Obviously you would need a standard, but it could be simple.
b: I wonder what this would to to traffic flow if everyone was slowing down in anticipation of the lights turning green. The people who understood what has happening would be OK with it, but the unenlightened might honk a lot.
Posted by: mahonj | 25 August 2011 at 12:59 PM
The unenlightened often whiff a bit! ;-)
Posted by: Davemart | 25 August 2011 at 01:51 PM
Tests work because the test drivers are enlightened. In the real world, most people are unenlightened, speeding and tailgaiting all the time, and don't slow down for a red light until long after the point where going faster does no good.
Suppose you know exactly when the light will turn green. A lot of people see the yellow on the crossing street, and start speeding up before the light turns green. That way they can screeam through the intersection. But, that's dangerous because somebody could be running the red light. having .33 seconds precision from a computer to tell you when the light will change? That just makes it more dangerous. Most people don't look both ways before they go through a green light, epecially if they are stopped, first in line. I watch too many people just keep looking straight ahead when the light turns green and plow right through the intersection without a worry. What the MIT people need to do is to enable smart phone cameras to look both ways before the driver goes through the intersection.
Why can't these MIT and Princeton folks invent a traffic light that doesn't change when a car makes a right turn on red? That could save a lot of energy and frustration for the drivers on the crossing street.
Posted by: Zhukova | 25 August 2011 at 04:31 PM
The virtue of SignalGuru is that it promotes signal-timed driving even where the smart infrastructure hasn't been rolled out. When the signals are finally equipped to broadcast their schedules, the vehicles and drivers will already be ready to take advantage of the improved information.
A critical mass of SignalGuru-equipped cars will slow the rest of traffic to match. This will spread the fuel savings beyond the participating drivers. If the total fleet can save 10%, that's huge; it's nearly a million barrels per day for the USA alone.
Posted by: Engineer-Poet | 25 August 2011 at 04:54 PM
Zhukova is right - this is a wacko idea.
Just put big indicators on the lights.
When lights are simply timed, traffic flows better - but apparently "timed for which way, N, S, E or W?)" is an issue.
Just put big indicators on the lights.
Even then, half the drivers just use the steady moderate speed of traffic to weave thru and get to the lead.
Just put big indicators on the lights.
It is EZ to watch the pedestrian timer to know if you can make it before red, or should just slow down; but they are not on most lights, some time out but the light stays green and they do not tell you how long until green.
Just put big indicators on the lights.
Posted by: ToppaTom | 25 August 2011 at 06:38 PM
That's right, TT, I've seen a few traffic lights that offer a count-down number sign of so many seconds before the light will turn red, right next to the light itself. This is very convenient for drivers to decide whether to speed up, or to slow down.
My daily route does not offfer such, however, since I am so familiar with all the traffic lights on my daily route that I was able to traverse many traffic lights without requiring much braking nor stopping.
Pedestrian lights offfer a great clue: If pedestrian light is green, you can speed up even at long distance without problem. Pedestrian light turns red and flashing means to tell you to speed up only you're close to the intersection, otherwise better slow down if you are too far away and be prepared to stop at the intersection if you're far away.
Posted by: Roger Pham | 25 August 2011 at 11:15 PM
The new step would be to put the inefficient human driver in the back seat and let the on-board computer communicate with the signaling system and do the driving. More fuel efficiency and a lot less accidents.
Posted by: HarveyD | 26 August 2011 at 09:12 AM
I forgot, does would think that driving a vehicle is fun and essential to their well being could go to the local private race tracks and pay for the privilege.
Posted by: HarveyD | 26 August 2011 at 09:15 AM
Hypermiling in my 2007 Prius is just as fun a game for me now as when I played Galaxian or Galaga back then...even better, no sore thumbs...Trying to pass thru as many traffic lights without stopping is another challenging fun that can rival traffic light drag racing in my younger days! Driving is definitely fun!
Posted by: Roger Pham | 26 August 2011 at 01:45 PM
RP....more fun (to drive) is not the ideal objective. Getting from point A to Z safely, with less stress, in a shorter time, using less energy, creating less GHG etc should be the objectives.
All those who want to have fun at the wheel should use the local race tracks and pay the full cost, including breaking their neck (and others) etc.
Let's not mix useful-essential driving with fun useless driving.
Posted by: HarveyD | 27 August 2011 at 10:12 AM
I don't think RP is the mixed up one here.
Posted by: ToppaTom | 27 August 2011 at 11:32 AM
Harvey, sometimes I think you just don't like FUN. IMO FUN is a prime directive for people.
Posted by: Reel$$ | 27 August 2011 at 12:21 PM
It is not for me.
We had a young, immature driver with an ultra noisy high power FUN car spinning its wheels (and doing 360s) in our area (with 2000+ hi-rise apartments-condo at 03:30h for the last few nights. Many residents were almost ready to use their shot guns. It took 6+ police cars to catch him. Two others are doing the same with their Harley Davidsons and they are not appreciated either. They are harder to catch. A few oil patches could fix them.
All FUN driving should be done on isolated race tracks, at least 10 miles away from residences, NOT on city and town streets. Noise is one of the worse pollution created by ICE vehicles.
Posted by: HarveyD | 28 August 2011 at 08:22 AM
FUN does not have to be noise and donuts. Many good citizens consider driving a car on an open freeway, down a country road or an open stretch of desert - F U N.
Resident Canadians own shot guns?
Posted by: Reel$$ | 28 August 2011 at 11:36 PM
F U N Driving was and still is a man-made 'created' culture artificially promoted at great cost by the Big-3 and Oil Cos during the last 90+ years. The aim was to sell more and more lemon gas guzzlers. That is certainly not why roads and highways were built.
Unfortunately, many million of us strongly believed it and got long lasting brain washed with their repeated Adds. USA lived for FUN driving for decades as the majority used to live for FUN smoking. Both were ridiculous. De-programming will take time (specially for the last 20%) and could be very costly. A very deep extended recession (1929-1939 style) may help the de-programming efforts. Otherwise, it may take a few generations.
Posted by: HarveyD | 29 August 2011 at 09:28 AM
I certainly feel your sentiment. Many have been lead astray to a hedonistic lifestyle that may end up in self-destruction (music and movie stars).
However, the latest HEV's from Audi, Toyota, Ford, GM, Huyndai, Honda, etc. have shown that the most fuel-efficient cars can also accelerate very fast and be fun to drive as well. The 2012 Camry Hybrid can go from 0-60 in 7.7 seconds, wow, that's Mustang GT 5.0's performance at one time, or even earlier model Corvette! While capable of 43 mpg fuel efficiency and can carry 5 passengers in roomy comfort!
Through technological advancement, we can have it all!
With gradual deployment of renewable energy and environmental preservation, we will be able to enjoy very high living standard in the future while enabling high employment rate and social stability.
Posted by: Roger Pham | 29 August 2011 at 11:32 PM