Building and infrastructure inspection methodologies need to be augmented with technology such as AI to help reduce errors.
The Signal Through The Noise: Geospatial Data Overload
How to best manage the overload of different types of data in engineering, architecture and building management.
Data. Everyone’s new best friend. The real estate and AEC space has been inundated with data lately. Throughout the entire lifecycle of a built asset we are now producing more data than ever. From financing and development to engineering architecture and construction to building operations and management.
Systems like CAD, BIM and more have made our buildings and infrastructure data producing machines. This includes design data, drawings, geospatial, sensor data, tenant/occupant information, access control logs, etc. This represents hundreds of different file types stored on different systems (both local and in the cloud) with different stakeholders and with different update frequencies (from data being produced multiple times per second to once every few years.
Even in an individual category such as geospatial inspection data contains separate types. Building inspections can be photos, video, 360-degree photos, LiDAR point clouds, orthomosaics or photogrammetry. Photos and video can be in varying spectrums including RGB, infrared and multispectral.
Not to mention that each of these types can come in various file formats with varying levels of compression. Geospatial data also tends to take up significant amounts of space. High-resolution 3D models or orthomosaics of a medium-sized building or bridge can easily be multiple gigabytes. This is compared to sensor data in simple text file format.
Not only is sensor data significantly smaller in size, it is also easier to analyze when each datapoint can have multiple attributes, including timestamps. Geospatial data is different. Inspection photos rarely have tags or other datapoint attached to them, let alone corresponding annotations referring to specific components of those photos (now made possible in platforms like T2D2 through standard open JSON overlays). While geospatial data may often have GPS and elevation coordinates, they are often not reliable at the accuracy required for real estate (ask us about the T2D2 data capture guide to improve your GPS accuracy).
Centralizing your inspection data onto one visualizer (even if the data itself lives elsewhere), enabling metadata search, and improving GPS accuracy are some of the first steps that you can take to extract more value out of building condition images.