Research

Framework for semantic integration simulation model

Team Mind Computation Lab
Investigator Professor Seung Wan Hong, Inha University
Collaborative Project associated with Yonsei University
Related publication Hong, S.W., Kim, H., Lee, J., Song, Y., & Yang, H. (2022). Development on an Intelligent Behavioural Simulation Model Based on Human Object Database. Korean Institute of Building Information Modeling Annual Conference 2022: BIM Culture-BIM Practices Towards a New Normal (pp.85-86).
Funding details This project was supported by the Korea Agency for Infrastructure Technology Advancement(KAIA) grant funded by the Ministry of Land, Infrastructure and Transport (Grant Number 22AATD-C163269-02).

Architectural design needs to deal with both general and specific user behaviors, yet predicting human factors in any given design and assessing its performance can be challenging. To address this, agent-based simulations have been studied as an inductive method to compute social interactions and spatial usage among users within physical constraints. However, the complexity of generating and analyzing user behaviors and programming their behavior rules has limited the use of these simulations among architectural professionals.

This project introduces a simulation framework that automatically generates user behaviors by integrating Building Information Modeling (BIM) data with Human Objects data. This project has established a framework for the semantic integration simulation model of human-object behavior and spatial data based on the theoretical model, 'Form-Function-Use' proposed by Kalay et al. (2014). Building on the theoretical model, the framework is structured with three key intelligent human object data: (1) Human Objects Data, (2) Activity Data, and (3) Behavior Rules Data and set data modules to integrate the semantics coming from BIM and intelligent human object data. By enabling the automated use of spaces by simulated human entities, this framework aims to provide architects to analyze human factors such as spatial bottlenecks and occupancy rates in design praxis.

F-F-U model (Kalay, Schaumann, Hong, and Simeone, 2014)

Database Structure

We firstly validated the framework of data structure and computability of the proposed simulator via several mockup simulations of a train station. Our research team collected occupants' authentic behaviors in public facilities (e.g., train stations), then we defined the data set to the following categories : (1) human objects consisting of physical and semantic properties, (2) activity, including motion data, (3) behavioral rules, responding to types of the built environment.

Intelligent Human Object Data Structure Overview

The data structure was refined to apply to residential buildings with three modules: (1) Human Objects Data: This includes both the physical and semantic properties of humans, where the semantic aspects are defined based on the user's occupational group and distribution, facilitating the implementation of social interactions. (2) Activity Data Module: This consists of unit behaviors linked with semantic data, where each unit behavior is associated with motion data enriched with semantic attributes. (3) Behavior Rules Data: This module combines Human Objects, Activity Data, and Building data into complex computations to produce simulations that allow human objects to interact with given spaces.

Data Modules for Integrating BIM and Human Object Data

Based on two data modules, virtual-users' behaviors are compiled and are ready to respond to the geometrical and semantic objects of BIM. We confirmed that the simulation could receive the building data by extracting BIM with geometric colliders and integrating it with intelligent human object data in the simulation model, providing quantitative data and qualitative design performances.

MODULE 1 Human object data input - Simulator Integration MODULE 2 Building data (BIM input) & Simulator Integration

Contact

Mind Computation Lab
Directed by Professor Seung Wan Hong
1505 Hi-tech Center, Inha University, 100, Inha-ro, Incheon, Republic of Korea www.mindcomputationlab.com

Google Sites

Report abuse