The dynamics of agent-based models and systems provides a framework to face complex issues related to the management of future cities, such as transportation and mobility. Once validated against empirical data, the use of agent-based simulations allows to envision and analyse complex phenomena, not directly accessible from the real world, in a predictive and explanatory scheme. In this paper, we apply this paradigm by proposing an agent-based simulation system focused on pedestrian/vehicle interactions at non-signalized intersections. The model has been designed based on the results gathered by means of an observation, executed at a non-signalized intersection characterized by a relevant number of pedestrian-car accidents in the past years. Manual video-tracking analyses showed that the interactions between pedestrians and vehicles at the zebra cross are generally composed of three phases: (i) the pedestrian freely walks on the side-walk approaching the zebra; (ii) at the proximity of the curb, he/she slows down to evaluate the safety gap from approaching cars to cross, possibly yielding to let the car pass (appraising); (iii) the pedestrian starts crossing. The overall heterogeneous system is composed of two types of agents (i.e. vehicle and pedestrian agents), defining the subjects of the interactions under investigation. The system is used to reproduce the observed traffic conditions and analyse the potential effects of overloading the system on comfort and safety of road users.

Bandini, S., Crociani, L., Feliciani, C., Gorrini, A., Vizzari, G. (2017). Collision avoidance dynamics among heterogeneous agents: The case of pedestrian/vehicle interactions. In AI*IA 2017 Advances in Artificial Intelligence. XVIth International Conference of the Italian Association for Artificial Intelligence, Bari, Italy, November 14-17, 2017, Proceedings (pp.44-57). Springer Verlag [10.1007/978-3-319-70169-1_4].

Collision avoidance dynamics among heterogeneous agents: The case of pedestrian/vehicle interactions

Bandini, S;Crociani, L
;
Gorrini, A;Vizzari, G
2017

Abstract

The dynamics of agent-based models and systems provides a framework to face complex issues related to the management of future cities, such as transportation and mobility. Once validated against empirical data, the use of agent-based simulations allows to envision and analyse complex phenomena, not directly accessible from the real world, in a predictive and explanatory scheme. In this paper, we apply this paradigm by proposing an agent-based simulation system focused on pedestrian/vehicle interactions at non-signalized intersections. The model has been designed based on the results gathered by means of an observation, executed at a non-signalized intersection characterized by a relevant number of pedestrian-car accidents in the past years. Manual video-tracking analyses showed that the interactions between pedestrians and vehicles at the zebra cross are generally composed of three phases: (i) the pedestrian freely walks on the side-walk approaching the zebra; (ii) at the proximity of the curb, he/she slows down to evaluate the safety gap from approaching cars to cross, possibly yielding to let the car pass (appraising); (iii) the pedestrian starts crossing. The overall heterogeneous system is composed of two types of agents (i.e. vehicle and pedestrian agents), defining the subjects of the interactions under investigation. The system is used to reproduce the observed traffic conditions and analyse the potential effects of overloading the system on comfort and safety of road users.
paper
Agent-based modelling; Collision avoidance; Simulation
English
16th International Conference on Italian Association for Artificial Intelligence, AI*IA 2017
2017
Esposito, F; Basili, R; Ferilli, S; Lisi, FA
AI*IA 2017 Advances in Artificial Intelligence. XVIth International Conference of the Italian Association for Artificial Intelligence, Bari, Italy, November 14-17, 2017, Proceedings
9783319701684
2017
10640
44
57
http://springerlink.com/content/0302-9743/copyright/2005/
none
Bandini, S., Crociani, L., Feliciani, C., Gorrini, A., Vizzari, G. (2017). Collision avoidance dynamics among heterogeneous agents: The case of pedestrian/vehicle interactions. In AI*IA 2017 Advances in Artificial Intelligence. XVIth International Conference of the Italian Association for Artificial Intelligence, Bari, Italy, November 14-17, 2017, Proceedings (pp.44-57). Springer Verlag [10.1007/978-3-319-70169-1_4].
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/10281/186140
Citazioni
  • Scopus 6
  • ???jsp.display-item.citation.isi??? 5
Social impact