Toward teaching agent-based simulation

Prospective Health Technology Assessment allows early decision making for innovative health care technologies. The main idea is to combine available domain knowledge with advanced simulation techniques in order to predict the effects of medical products and to find bottlenecks and weaknesses within the health system. In our recent publications a hybrid simulation approach with System Dynamics and Agent-Based Modeling has been presented. Hospital workflows have been modeled by state charts within agent behavioral models and have to be instantiated each time an agent is entering a hospital. This paper presents a mechanism to generate agents dynamically from SD models and extends the previously presented hybrid approach by process-oriented Discrete Event Simulation for hospital modeling. It connects processes to health care institutions and not to persons traversing them. Two extended example case studies show potentials for medical decision making using the three simulation paradigms in a common environment.

Simulation of Product-Service-Systems Piloting with Agent-Based Models (Outlined Revision)

Background and Purpose: Agent-based modelling and simulation (ABS) is growing in many areas like, e.g., management, social and computer sciences. However, the similar trend does not seem to occur within the field of business process management (BPM), even though simulation approaches like discrete event simulation or system dynamics are well established and widely used. Thus, in our paper we investigate the advantages and disadvantages of agent-based modelling and simulation in the field of BPM in simulation experiments. Design/Methodology/Approach: In our research, we investigate if there is a necessity for ABS in the field of BPM with our own simulation experiments to compare traditional and ABS models. For this purpose, we use simulation framework MAREA, which is a simulation environment with integrated ERP system. Our model is a complex system of a trading company selling computer cables. For the verification of our model, we use automated process discovery techniques. Results: In our simulations, we investigated the impact of changes in resources’ behavior on the outcome of company’s order to cash process (O2C). Simulations experiments demonstrated that even small changes might have statistically significant effect on outcomes of the processes and decisions based on such outcomes. Simulation experiments also demonstrated that the impact of randomly distributed fluctuations of well-being have a diminishing tendency with the increasing number of sales representatives involved in the process. Conclusions: Our research revealed several advantages and disadvantages of using ABS in business process modelling. However, as we show, many of them were at least partially addressed in the recent years. Thus, we believe that ABS will get more attention in the field of BPM similarly to other fields like, e.g., social sciences. We suggested areas in BPM simulations, e.g., modelling of resources, be it human or technological resources, where there is a need for ABS.

Agent-based modeling and simulation have evolved as a powerful tool for modeling complex systems, offering insights into emergent behaviors and interactions among diverse agents. Recently, integrating large language models into agent-based modeling and simulation presents a promising avenue for enhancing simulation capabilities. This paper surveys the landscape of utilizing large language models in agent-based modeling and simulation, discussing their challenges and promising future directions. In this survey, since this is an interdisciplinary field, we first introduce the background of agent-based modeling and simulation and large language model-empowered agents. We then discuss the motivation for applying large language models to agent-based simulation and systematically analyze the challenges in environment perception, human alignment, action generation, and evaluation. Most importantly, we provide a comprehensive overview of the recent works of large language model-empowered agent-based modeling and simulation in multiple scenarios, which can be divided into four domains: cyber, physical, social, and hybrid, covering simulation of both real-world and virtual environments, and how these works address the above challenges. Finally, since this area is new and quickly evolving, we discuss the open problems and promising future directions. We summarize the representative papers along with their code repositories in https://github.com/tsinghua-fib-lab/LLM-Agent-Based-Modeling-and-Simulation.

Test-driven simulation modelling: A case study using agent-based maritime search-operation simulation

For about a decade, we recognize a growing interest in hybrid simulation. Mixing different modeling paradigms promises to overcome limitations inherent to single paradigm approaches and help build models that are both realistic and accessible at the same time. However, a commonly accepted formalism to specify this kind of model on the conceptual level is barely found. We introduce a hybrid agent-based modeling formalism that integrates elements from the most widely applied paradigms: agent-based modeling and simulation, discrete event simulation, and system dynamics. The so-called HABM formalism is founded on hybrid automaton theory and DEV&DESS. It supports, among others, the specification of dynamic structure agent-based networks. We apply HABM to a strategic workforce planning problem by modeling an organization as a complex adaptive system.

Autonomous vehicles (AVs) may have a transformative influence on transportation systems. AV use is expected to reduce crash, transport costs, and land use. However, the net effects of the autonomous vehicle are unknown because there are no fully-autonomous vehicles in reality. The simulation systems are indispensable to be built up not only for cost but also for obtaining massive data. An agent-based simulation is capable of simulating the entities in a complex system, such as the AV system. Considering the potential of agent-based models to complement existing models and the difficulty to understand, to duplicate and to compare different agent-based models, this paper review existing agent-based modeling of autonomous vehicles with the help of ODD protocol, which has been applied in the review of agent-based land-use change models and agricultural policy evaluation. The significant variables that researchers have taken into consideration in sensitivity analysis and in the different scenarios are explored in the paper. Fleet size is the primary variable researchers considered, and we also discussed the probable causes of different AV replacement rates. We have assessed the quality of reviewed articles to gain more acceptance from the agent-based modeling community. Most of the reviewed papers follow the ODD protocol of agent decision making. However, in order to improve the level of modeling transparency, the executable and related data shall be available, at least the data source and the programming language/simulation platform should be provided. Several recommendations for further research are also presented.

Agent-based modeling and simulation in architecture

There is little research concerning comparisons and combination of System Dynamics Simulation (SDS) and Agent Based Simulation (ABS). ABS is a paradigm used in many levels of abstraction, including those levels covered by SDS. We believe that the establishment of frameworks for the choice between these two simulation approaches would contribute to the simulation research. Hence, our work aims for the establishment of directions for the choice between SDS and ABS approaches for immune system-related problems. Previously, we compared the use of ABS and SDS for modelling agents' behaviourin an environment with no movement or interactions between these agents. We concluded that for these types of agents it is preferable to use SDS, as it takes up less computational resources and produces the same results as those obtained by the ABS model. In order to move this research forward, our next research question is: if we introduce interactions between these agents will SDS still be the most appropriate paradigm to be used? To answer this question for immune system simulation problems, we will use, as case studies, models involving interactions between tumour cells and immune effector cells. Experiments show that there are cases where SDS and ABS can not be used interchangeably, and therefore, their comparison is not straightforward.

Agent-based modeling and simulation is a way to model the behavior of populations of components and their interactions within a system. The key of this approach is to model the components of the system as autonomous agents and to simulate their behavior for evaluating the system as a whole. That is very useful for observing the emergence of properties in social, biological, environmental or financial systems, among others. Artificial immune systems, which is a sub field of artificial intelligence, comprises systems modeled by simplifying models from the biological immune system. If agent-based modeling and simulation is used as a laboratory for understanding the biological immune system then, it can also be used for transferring the observed principles into artificial immune system models or for evaluating models that have been already adapted for solving technical problems. This paper presents first, a methodology for transferring principles of the biological immune system into the field of artificial immune systems. Then, it presents a brief explanation of the behavior of the cells of the biological immune system, which are treated as internal agents inside a biological organism. Afterwards, the modeling of some selected type of cells and the simulation of the whole simplified system are presented. In the end, the principles of that simplified system are transferred into the design of an alarm management system for the smart grid.

Applications of agent-based modelling and simulation in the agri-food supply chains

Agent-Based Simulation (ABS), an interdisciplinary area embracing both the computer science and the social science, has attracted much attention and aided the understanding of socially complex phenomena. A current important issue in this research area is how to improve ABS effectiveness and comprehension, which makes further mutual influence between the computer science and the social sciences indispensable - e.g., (1) agent modeling involving learning mechanisms in the computer science and (2) social dynamics analysis needed in the social science. Such integration of these two areas would help fulfill the great potential of ABS, first in solving complex engineering problems using agent-based technology and second in developing and testing new theories on socially complex systems. This special issue features ABS papers from both of these important areas exploring new trends in ABS. The 10 papers composing this special issue start with papers by Nobutada Fujii and Hiroyasu Inoue analyzing the relationship between the network structure and system dynamics. In these papers, an agent-based computational economics approach has been active in applying agent-based technologies to financial and economic systems. Papers by Biliana Alexandrova-Kabadjova, Isamu Okada, TomokoOhi, and Nariaki Nishino cover consumer and financial markets using agent-based models. They test economic theory and examine market phenomena for market design. Agent-based simulation is increasingly used in application fields in the social sciences. Papers by Kiyoshi Izumi, Hideki Fujii, Hiromitsu Hattori, and Shigeo Sagai propose solutions for actual social problems such as injury prevention, traffic, and electrical power. Models are created based on behavior data, and the integration of an agent-based model and real data is a hot topic in this area. As the beginning of these technical papers, this issue starts by a position paper to give an ABS overview for understanding important issues in ABS from an overall viewpoint and for understanding state-of-the-art ABS. The information presented is invaluable in helping readers grasp the important features of ABS.

Models to understand the impact of management practices on retail performance are often simplistic and assume low levels of noise and linearity. Of course, in real life, retail operations are dynamic, nonlinear and complex. To overcome these limitations, we investigate discrete-event and agent-based modelling and simulation approaches. The joint application of both approaches allows us to develop simulation models that are heterogeneous and more life-like, though poses a new research question: When developing such simulation models one still has to abstract from the real world, however, ideally in such a way that the ‘essence’ of the system is still captured. The question is how much detail is needed to capture this essence, as simulation models can be developed at different levels of abstraction. In the literature the appropriate level of abstraction for a particular case study is often more of an art than a science. In this paper, we aim to study this question more systematically by using a retail branch simulation model to investigate which level of model accuracy obtains meaningful results for practitioners. Our results show the effects of adding different levels of detail and we conclude that this type of study is very valuable to gain insight into what is really important in a model.

Agent-based modeling and simulation is a simulation paradigm that allows focusing on individuals, their interactions, and the complex behavior that emerges from them. Agent-based simulations are typically developed in simulation platforms that provide features related to agents. One such platform is NetLogo, to which a reinforcement learning extension was made available recently. The extension provides commands for using the Q-Learning algorithm, but no evaluation on whether it simplifies the development of simulations is available. This paper presents a quantitative evaluation on using the extension in two simulations: the classic cliff walking problem; and a real-world, adaptive traffic signal control (ATSC) simulation. Results show that the size of simulations source code developed using the extension is smaller than those developed without using it, giving evidence that the extension simplifies the development of simulations

This article presents an overview of the agent-based modeling and simulation approach and its recent developments in transport fields, with the purpose of discovering the advantages and gaps and encouraging more valuable investigations and applications of agent-based models. We clarify the agent-based model from agents, the background of development, and the basic structure applied in transport systems. Then, the agent-based transport modeling toolkits are discussed. The applications of agent-based models in transport systems are reviewed in three time scale models followed by an additional discussion of hybrid modeling approaches. The extensive modeling of the beliefs, desires, learning, and adaptability of individuals and the optimization problems using agent-based models are explored. Besides, we point out some limitations in terms of calibration and validation procedure, agents’ behavior modeling, and computing efficiency. In conclusion, some recommendations are given and suggest potential and insightful directions such as Big Data and Digital Twin for future research.