Ruprecht-Karls-Universität Heidelberg
Optimization in Robotics and Biomechanics

Swarm-like behaviour in crowds of motocycles in Asian metropoles

Benjamin RehKatja Mombaur - IWR

For every foreigner it is an amazing experience to see the traffic flow of immense crowds of motorbikes in Hanoi and some other Asian cities where cars have not yet become the dominant means of individual transportation. There are hundreds of motorbikes crossing at each intersection, often without any traffic lights or signs, moving steadily forward, almost automatically avoiding obstacles such as other motorbikes and crossing pedestrians, and all this with surprisingly few accidents. It is not yet understood how the intuitive individual behavior of Hanoi motorbikers, results as a whole in a high-performance collision avoidance system with such a high throughput. However, it seems important to gain this understanding to improve the interactions of other complicated traffic flow situations (e.g. of cars in dense downtown areas or in constructions sites) or of crowds of pedestrians (e.g. in case of large events or crowded public train stations) by appropriate planning, guidance and instructions.

The goal of this project is to gain this understanding using a combined approach of

(1) experimental data acquisition and analysis

(2) Scientific Computing techniques: modeling the behavior of motorbikers based on optimization and learning, and performing virtual simulations,

(3) validating these models in the small but real world of motorbike-like swarm robots and at the same time creating a new way of controlling such robots.

This project is financed by the FRONTIER research fund of the University of Heidelberg within the third pillar of the Excellence Initiative.

K. Mombaur, orb@uni-hd.de
Last Update: 14.03.2017 - 11:06

The photographs in the header of this webpage have been taken at the Musee de l'Automate in Souillac, France