German website

project term: 2012

software project in the master's course preparing the thesis

participant: Alexander Bertram

With this work, Alexander Bertram won the 1. prize in the Univention graduation award presented at the Berlin Linux Day 2013

In order to prove the feasibility for dynamic traffic navigation, an ant algorithm had to be implemented in a real road map. This project was conducted in the research field traffic navigation.

The map data to be used was the open source map OpenStreetMap which complies to the wiki principle.

Note: The implementation linked here does not work anymore, because the interfaces of Open Street Map had been changed. But in a successor project, the interfaces have already been repaired and wait to be linked here. By mid October 2015, you will find a working version here.

A prototype had been implemented in Scala and published under the Apache open source license. The current version of software may be obtained here. A hardware with CPU QuadCore and 2,5 GB memory is recommended.

The prototype may be installed on any computer with a JavaVirtualMachine 1.7. A local version of the prototype is here. The software may be used with a browser using the local port 8080. The display of maps requires an open internet connection.

The currently provided version offers several maps extracted from OpenStreetMap referring to the greater Hamburg area with streets of different categories baring up to 150 nodes.

The current functionality enables a human user to change the feasible speed on any road segment and, thus, simulate traffic congestions. At any time, any nodes may be set for start and destination, and always the currently recommended route with the best travel time will be displayed.

An interesting feature of ant algorithms is that they toggle between several preferred routes stochastically if these routes do not differ by quality substantially. This enables them to be used as a uniform traffic distributor.

If this raised your interest: There are various opportunities to continue this work. Besides enhancing the software in the course of follow-up projects there is multiple need for conceptual improvements which may be used for graduation theses.