Use this URL to cite or link to this record in EThOS:
Title: Human behaviour in office environments : finding patterns of activity and spatial configuration in large workplace datasets
Author: Koutsolampros, Petros
ISNI:       0000 0005 0289 5757
Awarding Body: UCL (University College London)
Current Institution: University College London (University of London)
Date of Award: 2021
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
The study of human behaviour in office spaces has a long and varied history from the 1970s to a recent resurgence of interest today, examining elements of collaboration and activity and how those elements are affected by the design and configuration of space. These studies however produced scattered and some times contradictory results, due to the lack of larger datasets and common sets of methodologies. This thesis examines one such large dataset using a newly developed unified framework that includes a common structure for all data, a spatial model to represent configuration in multiple scales and a set of statistical methods to extract meaningful information. The dataset contains around 40 companies in the UK, most of which are based around London and are of different scales, from single-floor workplaces to large multiple-campus offices. A complete workflow for working with this dataset is described, including existing metrics from Visibility Graph Analysis in extensive detail, but also newly developed ones such as 'Travel Concentration', a metric meant to capture attractor-driven effects. The analysis focuses on examining spatial configuration against movement and interaction in three scales, at the floor level (macro), the room level (meso) and the location level (micro), allowing for insights to emerge for the various parts of the design process. A variety of statistical models is presented with different levels of predictive strength, and which can be used for different purposes, but which may also depend on the size of the dataset and how biased an approach is to be taken. The results show that, in general, movement was more predictable than interaction, but also that the latter was a much more complex activity that became more predictable when broken down to other types, such as visiting and chatting interactions. More specifically, it was found that in the larger scales, both activities were mainly affected by the seat density of the workplaces, while in smaller scales the attractor-driven nature of movement became more apparent. Interaction on the other hand was found to relate very much to the availability of space and thus potential people to interact with as it happened mainly in workspaces. The thesis provides these results in the form of characteristics of spaces that tend to attract each activity (as predicted by each statistical model), but also as actionable insights that a designer might use in the design process.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available