On December 17, 2021, Obayashi Corporation, NTT DoCoMo (hereinafter, DoCoMo), and NTT Comware (hereinafter, NTT Comware) will utilize drones at indoor construction sites to efficiently manage the progress of construction work. Announced the success of the demonstration experiment.
Drone touring the construction siteIn the construction industry, many photographs are taken for recording and progress confirmation, such as situation photographs and progress photographs for each work process in construction, but in many cases it is "floor" because it is difficult to determine the position information inside the building. And the location information such as "construction zone" is manually input to the photo management system. Further, in order to compare photos of different dates, it is necessary to search for photos with the same location information from folders of each date, which is a problem that it takes time and effort.
In the progress management using drones carried out by the three companies this time, the drones automatically took pictures of the construction site multiple times at intervals of several days. You can view the construction photos taken each day by clicking on any part of the 3D data. This makes it easier to compare photos taken at the same location in chronological order, and it is possible to automate labor-intensive work, reducing the burden on the site.
In the demonstration experiment, we used Skydio's autonomous flight drone "Skydio 2" and NTT Comware's drone inspection software "Knowledge Map 4D", a drone flight program developed by DoCoMo for technical verification. From July 15th to August 3rd, 2021, the experiment was conducted in an indoor space of about 500 square meters at a construction site in Shinagawa-ku, Tokyo.
The drone, which remembered the locations of 116 locations, took pictures of the construction site a total of three times at intervals of up to 10 days, and confirmed that it was possible to accurately and autonomously fly the routes set in environments with different weather and time zones.
Skydio 2 has obstacle avoidance performance because it can estimate the surrounding three-dimensional environment and its own position from the images acquired by a total of six cameras mounted on the top and bottom of the aircraft. Because it is small and can fly even in narrow areas, it is possible to fly while shooting even in a complicated space such as a construction site. Furthermore, by using the flight program for technical verification developed by DoCoMo for Skydio 2, it is possible to preset routes that pass through multiple points, and even in environments where it is difficult to obtain location information such as GPS, complicated routes can be automatically set. Fly autonomously.
Flight trajectory and shooting point (KnowledgeMap 4D) generated from the image of the drone and placed in 3D spaceKnowledgeMap 4D can arrange the flight trajectory and shooting point of the drone in 3D space, so it is possible to accurately grasp the position and altitude of the drone at the time of shooting. By converting the captured images into 3D data and importing them into the software, it is possible to reflect the shooting locations of construction photographs in a 3D space that reproduces the construction site. As a result, it was confirmed that by selecting any point from the point cloud displayed on KnowledgeMap 4D, construction photos with different dates are displayed for each shooting location, and comparisons can be easily made in chronological order.
In the future, the three companies will continue to work together to further improve the efficiency of work at construction sites that utilize drones.
Comparison image of construction photos (left is July 15th, right is the situation of the same construction site on August 3rd)