Dynamic Carbon Activity Mapping in Urban Environments
Research Team Lead
James Elder
James Elder
York University
Accurate characterization of the spatial distribution and dynamics of carbon activity is very important for achieving sustainability in our use of energy resources and for the protection of our environment. Two key urban loci of energy use and emissions are buildings and vehicles. This project examines the use of visual and other forms of sensing to track human activity, traffic and heat loss from buildings in order to estimate energy use and emissions in urban environments. This in turn will help stakeholders to channel resources effectively when designing, constructing and managing more energy-efficient buildings and optimizing road networks and public transportation services, in order to reduce energy consumption and carbon emissions.
Research Description:
dynamic-carbon-activitiy The lack of reliable methods for continuous monitoring of carbon emissions and energy loss from individual buildings is a critical barrier to addressing this problem. Here we propose to develop a 3D GeoWeb service for continuous collection, distribution and visualization of the energy and emission activity of individual buildings. This will include information about energy consumption for heating, cooling and lighting, energy loss information, and information about human occupancy, obtained using both optical and thermal surveillance cameras, and distributed over wireless sensing networks. Data will be visualized at selectable levels of detail through 3D cityscape (indoor and outdoor) models based upon virtual indoor (BIM) and outdoor (GIS) models.

The research team will consist of experts in geomatics (Sohn), computer vision (Elder) and visual interfaces (OCAD), as well as industrial experts in environmental monitoring software technology. The research team at York University will build upon a prototype system, developed in earlier work, that allows 3D visualization of urban dynamics. The team will augment this system to incorporate continuous data on 1) heating, cooling and lighting operations, 2) building heat loss estimated with thermal camera technologies, 3) building and room occupancy estimated using person detection and tracking with visual surveillance cameras, and 4) vehicle tracking to estimate traffic flow. In this way, spatiotemporal information on carbon activity in relation to human activity will be analyzed, localized and visualized through 3D indoor/outdoor cityscape models. To achieve this goal, the research team will implement a dynamic GeoWeb system that manages 1) real-time data acquisition of wireless sensors distributed in indoor and outdoor environments (sensor layers); 2) data protocols to integrate SensorML, IndoorML and CityGML (protocol layer); 3) data management layers to handle indoor and outdoor 3D models and associate semantics, real-time sensor and moving object information (data layers); 4) dynamic response and on-demand rendering of tracked objects (response layers). Existing object detection and motion tracking algorithms will be improved to more reliably detect, classify and track people and vehicles using optical and thermal surveillance cameras.

A comprehensive interaction between the user and city models to retrieve carbon activity information will be key to the system's usability. A particular challenge is to render avatars that will effectively represent estimated occupancy and human activities. Also important will be the way in which the level and localization of carbon activity is represented, and how multi-floor indoor environments can be rendered in an intuitive and uncluttered fashion.
Gunho Sohn
Faculty Researcher, York University
Petros Faloutsos
Faculty Researcher, York University
Eduardo Del Soto
Research Assistant, York University
Emilio Almazan
Post-doctoral Research Fellow, York University
Eros Gulo
Master Student, GeoICT Lab, ESSE Department, York University
Hadi Hor
Ph.D. Student, GeoICT Lab, ESSE Department, York University
Mojgan Jadidi
Postdoctoral Researcher, GeoICT Lab, ESSE Department, York University
Nada Elasal
Research Assistant, York University
Phillip Robbins
Master Student, GeoICT Lab, ESSE Department, York University