Abstract—The construction of world models typically requires laborious survey and cataloguing of a building’s contents. OpenRoomMap attempts to crowd-source this information by allowing all building occupants to contribute to the model. Our initial deployment of OpenRoomMap has considered our office building that has a floor area of approximately 10000 square meters over 3 floors. Building occupants have mapped approximately 70% of the building and placed over 4000 objects. We show how the accuracy of placement depends on the physical restrictions that constrain the positions of objects and discuss the strategies we are using to encourage ongoing update and maintenance of the data. Keywords-context-aware, world-model, crowd-sourcing, VGI
Andrew Colin Rice, Oliver J. Woodman