Modelling bus-on-demand using SUMO and TraCI.
SUMOoD (SUMO on Demand) is a TraCI-based implementation of the simulation presented in [1,2].
SUMOoD v0.1 is based on a model of demand-responsive transportation developed by Kanchana Sakulchariyalert, Russell Thompson, John Haasz, Stephan Winter and Priyan Mendis in Delphi [1]. This software was developed further by John Haasz, John McDonald and Nicole Ronald [2]. In this software, vehicles use a variant of the DARP (Dial-a-Ride Problem) algorithm to pick up and drop off passengers at nodes in a network. While pre-booking rides is permitted by the software, it was used to experiment with ad-hoc (e.g., immediate pickups) only.
This model was replicated in SUMO using the TraCI Python interface. SUMOoD deals with ad-hoc requests from passengers wanting to travel between two locations. Passengers were restricted to nodes so that results of both models could be compared. In this case, it meant a node was represented by one outgoing link in SUMO, about 50 metres past the intersection.
For more information about the model and potential extensions, please see http://imod-au.info/sumood.
SUMOoD has been known to work with SUMO v0.20.0, however was tested more thoroughly with SUMO v0.18.0. The patches are taken from v0.20.0.
Changes are required to the SUMO source code as well as the TraCI Python interface: see the Patches folder.
The request input files are named sumo-<scenario>-<config>-<run>-people.csv, where scenario and config are strings, and run is a 2-digit integer. In my input files, the config is an integer-character pair, where the integer signifies the number of vehicles (i.e., 3, 5, 8, or 10) and the character signifies the demand ("S", "M", "L" -> small (20 requests), medium (40 requests), large (60 requests)).
Vehicles are specified in a normal SUMO .rou file; the vehicle type is "taxi".
The simulation is started with:
python drt.py <config> <run>
This work has been supported by the Australian Research Council (LP120200130).
[1] Thompson, R. G., Sakulchariyalert, K., Haasz, J., Winter, S. and Mendis, P. (2011), Determining the Viability of a Demand Responsive Transport System, published at http://imod-au.info/thompson11/.
[2] Ronald, N., Thompson, R. G., Haasz, J. and Winter, S. (2013), Determining the Viability of a Demand-Responsive Transport System under Varying Demand Scenarios. Proceedings of the 6th ACM SIGSPATIAL International Workshop on Computational Transportation Science, Orlando, Florida.