ways2go, 4. Call (2011)
Urban traffic control systems are primarily optimized for motorized traffic. Cyclists typically have stop more often and wait longer than cars at traffic lights. In BikeWave we develop new features for an existing smartphone app for cyclists. We introduce an innovative routing method which learns traffic signal programs by analyzing GPS trajectories. The second feature is a driving assistant that provides hints for an anticipatory driving style which allows cycling longer distances with the same energy spent due to less braking. BikeWave therefore promotes cycling as an elegant, efficient and environmentally friendly mode of transport.
State-of-the-art routers for motorized transport already use real-time traffic information to calculate travel times. Considering time- and location dependent travel times in routing for cyclists is difficult. The high variance of the desired speed of cyclists and the dependency of speed and traffic control related waiting times creates a bias in the estimation of realistic travel times. One reason is the optimization of urban traffic control systems for motorized traffic. A green wave for cars and fast cyclists can transform to a red wave for slower cyclists. To enable a realistic calculation of travel times the router has to know either the statistical relation of velocity and waiting times or the actual waiting times at the next junctions.
1. Development and application of methods for analyzing cyclists’ trajectories from microscopic traffic flow simulation and GPS positioning.
2. Modeling of travel time variability of any routes considering statistical dependencies in waiting time models.
3. Realistic, individual bicycle routing in urban areas including individual cycling speeds, level of service and statistical distribution of travel times.
4. Implementation of a driving assistant providing hints of driving speed and contextual information
The prototype offers routing that considers a tradeoff between travel time and attractiveness to create sets of alternative routes. The recommended route as well as the green wave assistant help to minimize waiting times at traffic lights and make the ride more comfortable. In addition, contextual information about saved energy and reduced travel time are provided to highlight the service's benefits and to make it attractive even for experienced cyclists. The use of OpenStreetMap, GPS trajectories and self-learning waiting time models makes the prototype location-independent.