Dynamic Traffic Control Yishai Goldstein & Nahum Nir Computer Science & Engineering Hebrew University
Present Traffic Situation
The Goals To minimize annoyances in the driving experience To help the driver reach his or her destination quickly
Methods for Achieving Goal Improve traffic lights Improve signs
Dynamic Traffic Lights
General Algorithm
Aggravation When a car does not drive smoothly it will “accumulate” aggravation This aggravation can be represented as some number Each car starts off with some aggravation
Traffic Light Algorithm The more aggravation there is at a stop light – the more likely the stop light is to turn green No one waits forever Emergency vehicles may be provided with a switch for “infinite aggravation”
Ant Colony Many non-intelligent individuals Intelligent System Pheromones - Aggravation
Inside The Algorithm Junctions, Cars and Ants Decision Making − Junctions − Ants
Simulation Scale, jams, and JAVA jams
Simulation Proof of the system
Dynamic Directional Signs
Signs Today Example : Problem : Different routes maybe preferable, depending on the present traffic Solution : Have signs change dynamically according to present traffic, just as our traffic lights.
Traffic Sign Algorithm Same infrastructure (Hardware) Internet routing Dijksra’s algorithm for shortest path Path measured by cars
Product Presentation Seeing the traffic signs in action
Map
User Interface Overview The user - driver, will be exposed to traffic lights and directional signs, just as a driver is today Drivers will not need to learn a new system They will not even need to be aware that the system had been changed
Problems & Solutions
Is Big Brother Watching? No information of specific vehicle is used (unlike cell-phones) Extra information can be added if needed
Four wheels good, two legs bad What about pedestrians? − When cars have a red light – people have a green light − Push buttons can be added at crosswalks Not all cars must participate for system to work
Competitors Present system in Jerusalem − Camera based − Requires human staff at all times − Predefined scenarios only Pelephone − General directions not specific − No handling of lights
Future Improvements Driver can request a specific route − Special screen inside car − Connection to PDAs Traffic report Statistics accumulator
Thank You
Technical Overview
Hardware Intersection Computer Sensors Traffic Lights Traffic Signs Network Car Transmitter
Transmitters One in each car Aggravation builds in one byte of temporary memory Connected to speedometer Internal clock No need for permanent memory