Advancing the Layered Approach to Agent-Based Crowd Simulation

Banerjee, Bikramjit; Abukmail, Ahmed; Kraemer, Landon

doi:10.1109/pads.2008.13

Cited by 41 publications

(25 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Although the use of a spatial database and a physics middleware for a social-force-driven crowd simulation is not new [1,19], to the best our knowledge, no existing crowd simulation has specifically exploited the uniqueness of airport traveller experience or used the layered AI approach in combination with a physics middleware and the social-forces approach as a backdrop for a real-time virtual crowd setting in a video game situation. We use the available games technology, and extend them to fulfil our aims and goals.…”

Section: Fig 1 London Gatwick Airport South Terminal Map Adapted Fmentioning

confidence: 99%

“…Several methods that face the challenges of real-time crowd simulation have been proposed [1,7,17,18]. One of the earliest applications of agent-based modelling was seen in Reynold's boid method [17], which has recently been extended onto the Sony Playstation 3 architecture to simulate up 15,000 agents at 60 frames-per-second [18].…”

Section: Crowd Simulationmentioning

confidence: 99%

“…CA crowds have been criticised for providing unrealistic crowd animation [15]. Bannerjee et al [1] uses a scalable layered intelligence technique and a Semi-Markov Decision Process for a crowd simulation. They demonstrate navigation behaviours in a hall-evacuation scenario and they evaluate the scalability and computational performance of their approach in highly crowded scenarios.…”

Section: Crowd Simulationmentioning

confidence: 99%

“…One major advantage of the spatial database approach is its scalability as further layers can be easily added. Banerjee et al [1] recently extended the spatial database to simulate an agent-based crowd simulation.…”

Section: Static and Dynamic Representationmentioning

confidence: 99%

“…The force (1) If (1) is applied, the agents try to avoid static obstacles defined in the physical calculated path stored in the dependent distance which lie within…”

Section: Static and Dynamic Representationmentioning

confidence: 99%

See 4 more Smart Citations

Towards Agent-Based Crowd Simulation in Airports Using Games Technology

Szymanezyk

Dickinson

Duckett

2011

Agent and Multi-Agent Systems: Technologies and Applications

View full text Add to dashboard Cite

Abstract. We adapt popular video games technology for an agent-based crowd simulation in an airport terminal. To achieve this, we investigate the unique traits of airports and implement a virtual crowd by exploiting a scalable layered intelligence technique in combination with physics middleware and a socialforces approach. Our experiments show that the framework runs at interactive frame-rate and evaluate the scalability with increasing number of agents demonstrating navigation behaviour.

show abstract

Section: Fig 1 London Gatwick Airport South Terminal Map Adapted Fmentioning

confidence: 99%

Section: Crowd Simulationmentioning

confidence: 99%

Section: Crowd Simulationmentioning

confidence: 99%

Section: Static and Dynamic Representationmentioning

confidence: 99%

“…The force (1) If (1) is applied, the agents try to avoid static obstacles defined in the physical calculated path stored in the dependent distance which lie within…”

Section: Static and Dynamic Representationmentioning

confidence: 99%

See 3 more Smart Citations

Towards Agent-Based Crowd Simulation in Airports Using Games Technology

Szymanezyk

Dickinson

Duckett

2011

Agent and Multi-Agent Systems: Technologies and Applications

View full text Add to dashboard Cite

show abstract

A Hybrid Reinforcement Learning and Cellular Automata Model for Crowd Simulation on the GPU

Ruiz

Hernández

2019

Communications in Computer and Information Science

View full text Add to dashboard Cite

A Multi-planar Approach to Encoding Story Worlds for Dynamic Narrative Generation

Tree

Malizia

2021

Human-Computer Interaction – INTERACT 2021

View full text Add to dashboard Cite

Tabletop Role-Playing Games transport the player to a fantastical fictional world. This world is inhabited by characters, players and objects. Prior to the commencement of play, the Dungeon Master prepares not only the rules of the world but an account of historical events and the current world state. While the rules provide the player with boundaries of play, the character sheets store the current player state, and the back story provides context for the happenings during the game session.Consider how often a story is told where, "In a land far away lived…". Rarely does a story rely on a universe where all existents are spontaneously spawned into being. Instead, vast historical worlds with cultures, histories and languages are created. This context provides the reader/viewer/player with an understanding of the world in which they find themselves. Players learn of warring factions, key artefacts, and are introduced to key drivers of the worlds drama. When Non-Playing Characters(NPC) are introduced into the game, the dungeon master draws on this history to form quests and to feed the players a view of this world.These behavioural systems, however, react to a player's moves with little consideration for the world context. This lack of context leads to the trope of the robotic NPC with little awareness of the world. How do we ingrain this knowledge of a contextual world? And even more critically at this seminal stage.How can game developers build, store and control dynamic story worlds?This question forms a component of the more significant question of how human authors can work alongside Artificially Intelligent systems to tell stories in games that adapt based on player action. Fundamental to this challenge is a unified model of story worlds that is comprehensible to human and machine alike. The necessity to model whole worlds is a challenge that has previously been undertaken in the field of crowd simulation, where we draw our inspiration.

show abstract

Advancing the Layered Approach to Agent-Based Crowd Simulation

Cited by 41 publications

References 7 publications

Towards Agent-Based Crowd Simulation in Airports Using Games Technology

Towards Agent-Based Crowd Simulation in Airports Using Games Technology

A Hybrid Reinforcement Learning and Cellular Automata Model for Crowd Simulation on the GPU

A Multi-planar Approach to Encoding Story Worlds for Dynamic Narrative Generation

Contact Info

Product

Resources

About