New version of “Tempus”, a framework for multimodal route planning

By Thursday December 18th, 2014Developpement, NewsFR

Oslandia and IFSTTAR are pleased to announce the release of a new version of their multimodal route planning framework: Tempus

Tempus is an Open Source C++/Python framework that allows to use, develop and test route planning algorithms with a particular focus on
multimodal routing and multi-objective requests, where every possible transport modes are taken into consideration for a trip: private cars, public transports, shared vehicules, bikes, etc.

Tempus features :

  • a C++ core based on Boost::graph for efficient and generic multimodal graph manipulations
  • a (dynamically loadable) plugin architecture for routing algorithms
  • exposure of its `API <http://ifsttar.github.io/Tempus/api/index.html>`_ as a webservice through a FastCGI WPS server that is able to respond to simultaneous requests
  • a client for QGIS
  • a loader script that is able to load road network data from Navteq, TomTom and OpenStreetMap, public transport data from GTFS and point of interests from CSV or shapefiles
  • a Python request API

Tempus is distributed under the terms of the GNU LGPLv2+

Tempus is on GitHub, get it, fork it, PR it : https://github.com/ifsttar/Tempus

The focus of Tempus is both on delivering an engine able to run efficient state-of-the-art routing algorithms and on offering a complete development environment to designers of routing algorithms.
A complete set of data from different sources can be integrated into a main (PostGIS) database thanks to the loader part. The modular architecture allows to develop algorithms as plugins for the
core. Some mechanisms are dedicated to help during the development: user variables for plugins, possibility to output parts of the traversed graph, user metrics, etc.

This 1.2.0 release brings:

  • a routing algorithm able to deal with **turning movement restrictions** for each transport mode
  • a reverse graph adaptor that allows to see the **graph reversed**. Requests with an ‘arrive before’ constraint can then be processed.
  • a tool (in the QGIS plugin) to select a **subset** of the graph to work on
  • a support for **speed profiles** on roads
  • an implementation of the A* algorithm

As for previous releases, a documentation is available, including notes for compilation and installation.A Windows installer is also available. It can be used to install Tempus on Windows, as well as (if necessary) QGIS 2.4 and PostGIS 2 !

Developments are still going on to include state-of-the-art algorithms, such as contraction hierarchies or Raptor.

Please let us know if you are interested in development, training or consulting around these pieces of technology. Do not hesitate to contact us at infos@oslandia.com .