Development of a 3d Indoor Model for Management of Emergency Evacuation in Public Buildings.Case Study of Ardhi House – Nairobi.

Abstract

Buildings have become a significant subject of recent disasters, either of natural or anthropogenic causes, resulting in massive casualties and damages in Kenya and almost everywhere else in the world. In the January 2019 travel advisory by the UK government, public buildings have been pointed out to be soft targets as a result of poor disaster preparedness and high human traffic. Emergency evacuation in buildings is particularly challenging because of the complexity of structures, poor situational awareness and presence of other hazards like gas leaks and falling structures. Indoor navigation is also much slower due to slow walking speed, the uncertainty of routes, blockages and numerous dead ends. The project aimed to design and develop a 3D indoor model incorporating mission-critical information for the management of indoor emergency rescue. The case study of this project was Third Floor, Wing C of Ardhi House. Indoor point cloud and panoramas of the observation area were obtained using Matterport Pro2 Lite 3D Camera MC250. Post-processing of the collected data was carried out through the Matterport cloud service using the Unity Engine plugin. The captured raw data was collected as a point cloud, depth data and panoramic images from the camera. The project identified critical datasets to include indoor building data, dynamic and semantic building information and outdoor emergency information. These datasets were then incorporated into the model. The resulting 3D model was then enhanced to obtain a navigable 3D web scene augmented with mission-critical data accessible via a web browser. The project evaluation was then undertaken by querying for the mission-critical data, navigation and orientation. The project’s deliverables were limited especially by the short time available to undertake the project, high cost of data collection, limited functionality of the Matterport Workshop and the inefficiencies of the camera such as heating up and alignment errors. The developed model was noted to be a step forward in availing critical information to rescue teams and victims during disasters and greatly improves situational awareness. The model achieved high accuracy in space representation, the camera was efficient in data collection and processing and the model was effective in information dissemination. The model could be improved to include all parts of buildings, include search and navigation functionalities and simulation of building collapse, fire and flood. The model may also be incorporated in other fields like cadastral mapping, urban planning, tourism, facility management and real estate valuation.