Dhirendra Singh

AAMAS Conference 2022 Conference Paper

Testing Requirements via User and System Stories in Agent Systems

• Sebastian Rodriguez
• John Thangarajah
• Michael Winikoff
• Dhirendra Singh

Agile software development is a popular and widely adopted practice due to its flexible and iterative nature that facilitates rapid prototyping. Recent work presented an agile approach to capturing requirements in agent systems via user and system stories. User and system stories present the requirements from the user and system perspective, respectively. Each story contains a set of acceptance criteria, which are a set of statements that identify the conditions under which the system behaviour can be accepted by the users or stakeholders. In this paper, we present a novel approach to testing the requirements that are specified via User and System stories in an agent system. We do this by developing a systematic approach to validating the execution traces output by the system against the specified acceptance criteria for each story. The approach identifies acceptance criteria that are met successfully in execution and those that fail. We present a fault model that categorizes the failures providing insight to the developers to address the failed cases. We classify three kinds of faults for a given acceptance criterion: (a) the trigger condition is never met; (b) when the trigger occurs the preconditions are not met; or (c) the trigger and preconditions are met but the resulting actions are not as expected. The motivating application of our work, which is also the test-bed for evaluation, is an agent-based simulation application for modelling the behaviours of civilians in a bushfire emergency scenario that is used in practice. We show our approach is able to successfully test and uncover requirements that were not met in this application.

IJCAI Conference 2017 Conference Paper

Emergency Evacuation Simulator (EES) - a Tool for Planning Community Evacuations in Australia

• Dhirendra Singh
• Lin Padgham

This work addresses the problem of encoding cognitive agents that are capable of complex reasoning beyond simple rules, within agent-based models (ABM). This is particularly important for social simulation where agents represent people. We provide a general solution to this problem through infrastructure that allows the integration of state-of-the-art Belief-Desire-Intention (BDI) and ABM systems. In this paper, we demonstrate how this infrastructure is being used to help emergency services in Australia plan for community evacuations.

AAMAS Conference 2017 Conference Paper

Integrating BDI Agents with Agent-Based Simulation Platforms: (JAAMAS Extended Abstract)

• Dhirendra Singh
• Lin Padgham
• Brian Logan

AAMAS Conference 2017 Conference Paper

The Importance of Modelling Realistic Human Behaviour When Planning Evacuation Schedules

• Chaminda Bulumulla
• Lin Padgham
• Dhirendra Singh
• Jeffrey Chan

JAAMAS Journal 2016 Journal Article

Integrating BDI Agents with Agent-Based Simulation Platforms

• Dhirendra Singh
• Lin Padgham
• Brian Logan

Abstract Agent-based models (ABMs) are increasingly being used for exploring and supporting decision making about social science scenarios involving modelling of human agents. However existing agent-based simulation platforms (e. g. , SWARM, Repast) provide limited support for the simulation of more complex cognitive agents required by such scenarios. We present a framework that allows Belief-Desire-Intention (BDI) cognitive agents to be embedded in an ABM system. Architecturally, this means that the “brains” of an agent can be modelled in the BDI system in the usual way, while the “body” exists in the ABM system. The architecture is flexible in that the ABM can still have non-BDI agents in the simulation, and the BDI-side can have agents that do not have a physical counterpart (such as an organisation). The framework addresses a key integration challenge of coupling event-based BDI systems, with time-stepped ABM systems. Our framework is modular and supports integration of off-the-shelf BDI systems with off-the-shelf ABM systems. The framework is Open Source, and all integrations and applications are available for use by the modelling community.

ECAI Conference 2014 Conference Paper

Integrating BDI Agents into a MATSim Simulation

• Lin Padgham
• Kai Nagel
• Dhirendra Singh
• Qingyu Chen

MATSim is a mature and powerful traffic simulator, used for large scale traffic simulations, primarily to assess likely results of various infrastructure or road network changes. More recently there has been work to extend MATSim to allow its use in applications requiring what has been referred to as "within day replanning". In the work described here we have coupled MATSim with a BDI (Belief Desire Intention) system to allow both more extensive modelling of the agent's decision making, as well as reactivity to environmental situations. The approach used allows for all agents to be "intelligent" or for some to be "intelligent"/reactive, while others operate according to plans that are static within a single day. The former is appropriate for simulations such as a bushfire evacuation, where all agents will be reacting to the changing environment. The latter is suited to introducing agents such as taxis into a standard MATSim simulation, as they cannot realistically have a predetermined plan, but must constantly respond to the current situation. We have prototype applications for both bushfire evacuation and taxis. By extending the capabilities of MATSim to allow agents to respond intelligently to changes in the environment, we facilitate the use of MATSim in a wide range of simulation applications. The work also opens the way for MATSim to be used alongside other simulation components, in a simulation integrating multiple components.

ECAI Conference 2014 Conference Paper

OpenSim: A framework for integrating agent-based models and simulation components

• Dhirendra Singh
• Lin Padgham

The growing use of agent-based modelling and simulation for complex systems analysis has led to the availability of numerous published models. However, reuse of existing models in new simulations, for studying new problems, is largely not attempted. This is mainly because there is no systematic way of integrating agent-based models, that deals with the nuances of complex interactions and overlaps in concepts between components, in the shared environment. In this paper we present an open source framework, called OpenSim, that allows such integrated simulations to be built in a modular way, by linking together agent-based and other models. OpenSim is designed to be easy to use, and we give examples of the kinds of simulations we have built with this framework.

IJCAI Conference 2011 Conference Paper

Integrating Learning into a BDI Agent for Environments with Changing Dynamics

• Dhirendra Singh
• Sebastian Sardina
• Lin Padgham
• Geoff James

We propose a framework that adds learning for improving plan selection in the popular BDI agent programming paradigm. In contrast with previous proposals, the approach given here is able to scale up well with the complexity of the agent's plan library. Technically, we develop a novel confidence measure which allows the agent to adjust its reliance on the learning dynamically, facilitating in principle infinitely many (re)learning phases. We demonstrate the benefits of the approach in an example controller for energy management.

AAMAS Conference 2010 Conference Paper

Learning Context Conditions for BDI Plan Selection

• Dhirendra Singh
• Sebastian Sardina
• Lin Padgham
• Stephane Airiau

An important drawback to the popular Belief, Desire, and Intentions (BDI) paradigm is that such systems include no element oflearning from experience. In particular, the so-called context conditions of plans, on which the whole model relies for plan selection, are restricted to be boolean formulas that are to be specified atdesign/implementation time. To address these limitations, we propose a novel BDI programming framework that, by suitably modeling context conditions as decision trees, allows agents to learn theprobability of success for plans based on previous execution experiences. By using a probabilistic plan selection function, the agentscan balance exploration and exploitation of their plans. We developand empirically investigate two extreme approaches to learning thenew context conditions and show that both can be advantageousin certain situations. Finally, we propose a generalization of theprobabilistic plan selection function that yields a middle-groundbetween the two extreme approaches, and which we thus argue isthe most flexible and simple approach.