Buildots, a construction technology startup, is developing an image recognition system that monitors every detail of an ongoing construction project and flags delays or errors automatically. According to MIT Technology Review, it is already being used by two of the biggest building firms in Europe.
Some companies currently have an AI site inspector that matches images taken on site against a digital plan of the building. However, the Buildots system uses video footage from GoPro cameras mounted on the hard hats of workers. When managers perform a site inspection, the camera on their head captures video footage of the project and uploads it to image recognition software, which compares the status of many thousands of objects on site with a digital replica of the building.
According to the article, the AI also uses video feed to track where the camera is in the building so it can identify the exact location of objects in each frame. The system can track the status of around 150,000 objects several times a week. It can also tell which state the object is in, from "not yet begun" to "fully installed."
The system also captures data without the need to walk the site with spreadsheets or schedules. When construction sites were shut down to all but the most essential workers, managers on several Buildots projects were able to keep tabs on progress remotely.
Source: MIT Technology Review