Source-linked AI summary
Prototyping Virtual Reality Serious Games for Building Earthquake Preparedness: The Auckland City Hospital Case Study
Ruggiero Lovreglio, Vicente Gonzalez, Zhenan Feng, Robert Amor, Michael Spearpoint, Jared Thomas, Margaret Trotter, Rafael Sacks
TL;DR
Earthquake evacuation training needs approaches that better capture occupant behaviour and emergency conditions than traditional methods, while VR serious games remain underexamined for this purpose. The paper discusses their design requirements through the Auckland City Hospital case study, proposing behavioural and training prototypes whose assessment was informed by 170 participants. It also identifies technological, interaction, and ethical boundaries that constrain realism and implementation.
Problem
VR serious games have not been extensively examined for earthquake preparedness, despite limitations in traditional training and the need to understand occupant behaviour.
Method
The paper theoretically discusses VR serious-game advantages, limitations, and design components, then illustrates them through two Auckland City Hospital prototypes.
Results
The case study presents practical development pipelines and design solutions for a Behavioural Prototype and a Training Prototype, with component perceptions assessed by 170 participants.
Takeaways & Limitations
VR serious games provide a possible way to investigate earthquake evacuation behaviour and train occupants in public buildings.
Takeaways & Limitations
VR serious-game development remains constrained by BIM conversion limitations, technological challenges, and limited NPC realism.
Abstract
from arXiv · showhide
Enhancing evacuee safety is a key factor in reducing the number of injuries and deaths that result from earthquakes. One way this can be achieved is by training occupants. Virtual Reality (VR) and Serious Games (SGs), represent novel techniques that may overcome the limitations of traditional training approaches. VR and SGs have been examined in the fire emergency context, however, their application to earthquake preparedness has not yet been extensively examined. We provide a theoretical discussion of the advantages and limitations of using VR SGs to investigate how building occupants behave during earthquake evacuations and to train building occupants to cope with such emergencies. We explore key design components for developing a VR SG framework: (a) what features constitute an earthquake event, (b) which building types can be selected and represented within the VR environment, (c) how damage to the building can be determined and represented, (d) how non-player characters (NPC) can be designed, and (e) what level of interaction there can be between NPC and the human participants. We illustrate the above by presenting the Auckland City Hospital, New Zealand as a case study, and propose a possible VR SG training tool to enhance earthquake preparedness in public buildings.
1. INTRODUCTION
Earthquake preparedness depends on predicting occupant behaviour and training people in recommended responses, but traditional drills may not reproduce earthquake realism or behaviour. The paper therefore examines VR serious games and develops two prototype concepts for behavioural investigation and training.
- Risk-reduction approaches: Earthquake risk reduction combines behavioural design with training occupants to prepare for and respond during and after earthquakes.Both approaches depend on improving understanding of occupant behaviour and preparedness.
- Limitations of traditional drills: Traditional evacuation drills may poorly reproduce earthquake conditions because occupants can perceive the threat, creating a substantial realism gap.Drills may also trigger behaviours different from those observed in real earthquake evacuations.
- VR and serious games: VR and serious games can represent earthquake threats and provide more realistic evacuation scenarios than conventional training approaches.The paper positions these technologies as potential ways to reduce the gap between drills and real evacuations.
- Paper objective: The paper discusses VR serious games for understanding earthquake occupant behaviour and teaching occupants how to cope during and after earthquakes.It focuses on both behavioural investigation and practical training.
- Prototype framework: Two prototypes are proposed: a Behavioural Prototype for investigating human behaviour and a Training Prototype for teaching recommended earthquake evacuation practice.The Training Prototype follows New Zealand Civil Defence guidance.
2. VIRTUAL REALITY AND SERIOUS GAMES
Serious games and VR offer immersive, controllable environments for earthquake training and behavioural research, but their benefits remain constrained by realism, technology, and ethical challenges. The paper highlights how these systems can collect behavioural data while requiring careful design and further evaluation.
- Serious-game concept: Serious games prioritize education within game experiences, while allowing entertainment and education to overlap.The paper uses this framing to position serious games as training tools rather than purely entertainment products.
- Prior applications: Existing earthquake serious games provide preparedness or evacuation training, but their low realism can limit engagement.Prior applications include precautions against earthquakes and knowledge of safe evacuation procedures.
- Prior VR applications: Only two VR serious-game applications were identified in the literature, including one that trained drop, cover, and hold without post-earthquake evacuation.This indicates limited coverage of earthquake preparedness within earlier VR applications.
- Training and experimentation: VR serious games can safely represent hazards such as debris and blocked exits while providing controlled evacuation scenarios.They can also support investigation of reactions and decisions under different damage and visibility conditions.
- Behavioural data: VR serious games can capture movement, viewing direction, attended objects, and verbal protocols that reveal decision-making and information seeking.These data extend beyond observations typically available in classic evacuation drills.
- VR constraints: Immersive VR remains limited by ghost experiences, restricted tracking areas, motion sickness, multisensory limitations, and controller-dependent interaction.These technological constraints affect how users perceive and navigate virtual environments.
- Motion sickness: Motion sickness cannot yet be fully prevented reliably, so VR navigation and display solutions must be selected case by case.Visual-vestibular conflicts and display lag contribute to sickness and postural instability.
3. AN EARTHQUAKE GAME ENVIRONMENT
The VR serious-game environment requires design choices about the earthquake event, building representation, damage, surrounding hazards, and evacuee interactions. These choices depend on whether the prototype investigates behaviour or trains recommended evacuation practices.
- Earthquake event: The earthquake event must be compressed into a gaming timeframe because its three seismic stages can last several months.The stages are foreshock, main shock, and aftershock.
- Earthquake event: Because immersive technologies cannot physically shake users, earthquake effects can be conveyed through building impacts, moving objects, and auditory feedback.Physical shaking may alternatively be added with shake tables, seat systems, or controller vibration, although shake tables impose laboratory location constraints.
- Building selection and representation: Existing buildings provide familiarity but require detailed real-world data, whereas hypothetical buildings simplify development but cannot test layout familiarity as a behavioural factor.Existing-building representations may use BIM models, which can contain large datasets and lose material parameters during game-engine conversion.
- Building selection and representation: The virtual environment may also need external building and surrounding-area hazards that occupants encounter before reaching designated safe places.The required extent and level of external-environment detail remain design questions.
- Damage representation: Building damage can be represented quantitatively through structural modelling and fragility functions or qualitatively by mimicking damage from videos and photographs.The quantitative approach predicts non-structural damage from relationships between structural damage and failure probability.
- NPC behaviour and interactions: NPCs or multiple human participants can model social influence, but their design should reflect whether the prototype studies observed behaviour or trains best practice.NPCs may follow guideline-recommended or unsafe actions, while multi-participant games are constrained by available HMDs and graphics performance.
- Training objective: Post-earthquake training should identify and correct appropriate and inappropriate behaviours during the main shock and subsequent evacuation.Designers can derive target behaviours from national or international guidelines and the evacuation plans of represented buildings.
4. CASE STUDY
The Auckland City Hospital case study implements Behavioural and Training Prototypes in a Unity-based VR environment, combining earthquake effects, building damage, NPCs, and constrained navigation. The prototypes use a hospital setting to investigate evacuation behaviour and support earthquake-preparedness training.
- Case study purpose: The Behavioural Prototype and Training Prototype apply VR serious-game concepts to investigate earthquake-evacuation behaviour and train users according to New Zealand Civil Defence guidelines.The case study was developed through collaboration among civil engineering, computer science, and social science researchers with public-organization advisors.
- Case study setting: A section of Auckland City Hospital was selected to represent a public building used by staff and visitors, while enabling comparison with inductions, leaflets, and seminars.The hospital context also addresses the difficulty of conducting traditional drills where vulnerable populations are present.
- Virtual environment: The fifth-floor public and administrative areas were reconstructed in Unity from BIM models, 2D DWG files, site visits, and photographs, then optimized toward 50 frames per second.Lightmapping and Occlusion Culling were used to optimize rendering for the VR application.
- Earthquake simulation and damage: Earthquake effects were simulated qualitatively by shaking the participant’s floor, applying virtual forces through Unity’s physics engine, and generating damage equivalent to VII-VIII Modified Mercalli intensity.Object responses depend on mass, friction, sliding, and stability; gravity can make objects fall or overturn.
- Earthquake simulation and damage: Computationally demanding earthquake dynamics were applied only around the participant, while impacts elsewhere were generated from a pre-simulated scenario and damaged walls and ceilings replaced undamaged components.A vibrating platform reproduced physical shaking during the VR experience and was activated by low frequencies of the earthquake noise.
- NPCs and navigation: NPCs populated the hospital environment as non-interactive or interactive characters, while the Training Prototype used wait points, a first-person controller, and selectable action panels.The wait-point design limits navigation and ecological validity but was accepted because the Training Prototype prioritizes learning outcomes over full interactivity.
5. DISCUSSION AND CONCLUSIONS
The paper identifies key design components and development challenges for VR serious games addressing earthquake emergencies, illustrated through two Auckland City Hospital prototypes. The case study covers earthquake simulation, building damage, NPCs, participant interactions, navigation, and participant assessment.
- The framework identifies earthquake features, building representation, damage representation, NPC behaviour, participant–NPC interactions, and behavioural and learning outcomes as core design components.
- The Auckland City Hospital case study develops two VR serious game prototypes for improving earthquake preparedness in a public building, using feedback from 170 participants.
- The proposed Unity-based earthquake simulation balances the computational cost of dynamic modelling against realism and applies building damage across the full virtual environment during shaking.
- Earthquake serious games require more complex pre-evacuation actions than fire games because safe evacuation involves tasks before leaving the building.
- NPCs support social-interaction training and behaviour investigation, while navigation design must address motion sickness and balance realism with the serious game’s goal.