Abstract
This paper explores the potentials of using agent-based modelling (ABM) in computer game technology to create a dynamic, playful, and interactive pedestrian simulation that can serve as a better tool of communication between designers and project stakeholders to co-create human-centred public urban spaces. The paper demonstrates how quantitative and qualitative observational data about pedestrian spatial behaviour derived from ethnographic fieldwork on Potsdamer Platz in Berlin (Germany) constitute a set of backbone data to construct agent's spatial behaviour in the simulation environment. Such data includes interviews, counting pedestrians, mapping and documenting lingering activities, ground floor analyses, as well as photographing and field notes.
The Unity3D game engine was used to construct and visualise the agent-based pedestrian simulation. The simulation model considers three types of agent's behaviour. 1) purposeful and goal-oriented behaviour, 2) browsing behaviour, and 3) social behaviour. The study presented focuses on creating agents that simulate different behaviour witnessed in the observational studies on the various social groups on Potsdamer Platz. The agents were thus programmed to have the ability to navigate and avoid obstacles, perform lingering activities, as well as to be able to adapt instantly to the different physical settings of the proposed design alternatives.
The findings of this paper highlight the potential of using ABM and game technology in the realm of landscape architecture and urban design. Using game technology helps to construct real-time interactive, immersive, and highly visual agent-based pedestrian simulation. Hence, rather than using static plan drawings and physical models alone, the simulation served as a dynamic tool to convey intricate behavioural patterns in the context of design team meetings and workshops with real estate owners and municipal planners. The paper also identifies the structure sets of spatial behaviour parameters required for interactive pedestrian simulation, as well as the types of physical settings and their influence on the agent’s spatial behaviour.
The Unity3D game engine was used to construct and visualise the agent-based pedestrian simulation. The simulation model considers three types of agent's behaviour. 1) purposeful and goal-oriented behaviour, 2) browsing behaviour, and 3) social behaviour. The study presented focuses on creating agents that simulate different behaviour witnessed in the observational studies on the various social groups on Potsdamer Platz. The agents were thus programmed to have the ability to navigate and avoid obstacles, perform lingering activities, as well as to be able to adapt instantly to the different physical settings of the proposed design alternatives.
The findings of this paper highlight the potential of using ABM and game technology in the realm of landscape architecture and urban design. Using game technology helps to construct real-time interactive, immersive, and highly visual agent-based pedestrian simulation. Hence, rather than using static plan drawings and physical models alone, the simulation served as a dynamic tool to convey intricate behavioural patterns in the context of design team meetings and workshops with real estate owners and municipal planners. The paper also identifies the structure sets of spatial behaviour parameters required for interactive pedestrian simulation, as well as the types of physical settings and their influence on the agent’s spatial behaviour.
Original language | English |
---|---|
Publication date | 2019 |
Number of pages | 1 |
Publication status | Published - 2019 |
Event | City Futures 2019 - EURA UAA - University College Dublin, Dublin, Ireland Duration: 20 Jun 2019 → 22 Jun 2019 http://cityfutures2019.com/ |
Conference
Conference | City Futures 2019 - EURA UAA |
---|---|
Location | University College Dublin |
Country/Territory | Ireland |
City | Dublin |
Period | 20/06/2019 → 22/06/2019 |
Internet address |
Keywords
- Urban planning
- Human-centred urban design
- evidence-based design
- pedestrian simulation
- agent-based modelling
- game engines
- Unity3D
- urban life