The BDI Driver in a Service City (Extended Abstract) Marco Lützenberger Nils Masuch Benjamin Hirsch Sebastian Ahrndt Axel Heßler Sahin Albayrak DAI-Labor, Technische Universität Berlin Ernst-Reuter-Platz 7 10587 Berlin, Germany firstname.lastname@dai-labor.de ABSTRACT Most traffic simulation frameworks move vehicles from some location A to some location B as the result of different equa- tions of motion or fluid dynamics. As it is, reality is much more complex because what actually happens on the road is not only determined by physics of motion, but also by the perception and attitudes of the drivers. In this work, we introduce an approach which considers a driver’s state of mind within large scale traffic simulations. For this purpose we describe a BDI based conceptualisation of a driver and extend common simulation topologies with service oriented concepts. Categories and Subject Descriptors I.2 [Computing Methodologies]: Distributed Artificial Intelligence—Intelligent agents ; I.6 [Simulation And Mod- eling]: Model Development General Terms Human Factors, Experimentation, Measurement Keywords BDI, Simulation techniques, tools and environments 1. INTRODUCTION Despite the wide range of available traffic simulation frame- works, most products share the fact that the vehicle simu- lation is done in a pure computational fashion. Usually, the simulated vehicles are moved from a location A to a loca- tion B as a result of equations of motion or fluid dynamics. As it is, reality is much more complex, because what actu- ally happens on the road is not only determined by physics of motion, but also by the perception and attitudes of the drivers. A driver with a high affinity for public transport for instance might change his means of transportation when confronted with a traffic jam near a metro station and avail- able parking. This aspect does not affect the driving process Cite as: The BDI Driver in a Service City (Extended Abstract), M. Lützenberger, N. Masuch, B. Hirsch, S. Ahrndt, A. Heßler and S. Albayrak, Proc. of 10th Int. Conf. on Autonomous Agents and Multia- gent Systems (AAMAS 2011), Tumer, Yolum, Sonenberg and Stone (eds.), May, 2–6, 2011, Taipei, Taiwan, pp. 1257-1258. Copyright c  2011, International Foundation for Autonomous Agents and Multiagent Systems (www.ifaamas.org). All rights reserved. per se, but influences the traffic situation a fortiori. Several approaches [1, 2, 3, 5], integrate stimuli-reaction principles and mimic individual driving styles by implementing cogni- tive abilities for the simulated vehicles. Yet, a more com- prehensive, “strategic” consideration is mostly missing. In this paper we outline an according approach. We start by explaining the model we have specified for the driver and emphasise additional requirements for the topology model which are necessary to make this approach work. 2. THE BDI DRIVER IN A SERVICE CITY For our purpose, we have to address two topics. First, we have to define a model for the environment which is able to influence the behaviour of a driver by certain stimuli. Next, we have to define the behavioural model for the driver, which is able to comprehend the stimuli of the environment and is able to generate the driver’s action. The main difference between our approach and related work is that a driver is able to perceive and interact with his topology by making use of certain Infrastructural Fea- tures which may support the driver in achieving his goals, or influence his strategy in doing so. We define the term as follows: An Infrastructural Feature can be everything which is able to fulfil a desire (or parts of it) of a person at a cer- tain location of an infrastructure. As an example, consider public transport. It provides a service at many places of an infrastructure and supports a person’s desire to reach a certain location. Another example is a car park. Located at some location they provide service for any driver who wants to park his vehicle. According to our definition, In- frastructural Features are not necessarily related to traffic, but can also be interpreted as: Shop, restaurant, takeaway, telephone booth and many more. Based on our definition, it is nearly impossible to provide a complete model for any larger city; this is not our intention. Our objective is to provide a uniform way for the specification of these features in order allow for easy, custom definitions. We choose the Service Metaphor for this purpose and allow for a unified specification in terms of preconditions, effects, a scope, a lo- cation (or more than one, in case of a cross-linked service, such as a metro system) and a duration function. For the implementation of the Driver Model, we apply an agent oriented view [6] and follow a popular model for the conceptualisation of human behaviour: The BDI model [4]. This approach provides us with a specification for our im- 1257