At SECOM, when making a security plan or dealing with alarm signals, judgment is conducted by referring to an architectural drawing filled with lines and symbols to interpret the meaning and situation. However, to scale efficiently as the number of cases increase, machines must be able to make judgements by understanding the meanings and situation. To do so, it is important that machine-readable spatial data be created and machines can access it without human assistance. Machine-readable refers to the ease of a computer to process data. Data becomes machine-readable by including attribute information (metadata) which can be processed by a computer. Recently, 3D building model data in BIM (Building Information Modeling) is becoming popular.
In BIM, 3D building model data is expressed as an aggregation of building element objects. Thus data about walls, columns, and space can be handled as programmable attribute information.
However, when service providers such as SECOM use model data, two problems are exist. First, it is hard to extract only the partial data necessary for our service. Second, if different service providers each make a BIM model, version management of model data would be difficult and may conflict with each other. To solve this problem, we are researching "IFC Model Server," a web application to manage model data conforming to IFC (Industry Foundation Classes) standard.
The IFC Model Server is web application which manages IFC model data using a database, allowing data to be shared and used on the internet. It allows the storage and update of IFC model data, search of partial model data, and converts the search results.
Users can upload their IFC model data to the IFC Model Server. The IFC Model Server provides APIs for access to the stored model data. By calling different API functions, an user can extract partial data and convert it.
By providing these functions for users on internet, the IFC Model Server will realize a single point data access environment that allows users to use IFC 3D building model data effectively.