United States Department of Transportation
i
Rotorcraft Landing Sites – An AI-Based Identification System
-
2021-02-01
-
[PDF-7.54 MB]
Details:
-
Creators:
-
Corporate Creators:
-
Corporate Contributors:Rutgers University. Center for Advanced Infrastructure and Transportation ; United States. Department of Transportation. University Transportation Centers (UTC) Program ; United States. Department of Transportation. Office of the Assistant Secretary for Research and Technology ; United States. Department of Transportation. Federal Aviation Administration ; United States. Department of Transportation. Federal Highway Administration
-
Subject/TRT Terms:
-
Publication/ Report Number:
-
Resource Type:
-
Geographical Coverage:
-
Edition:Final report, 01/1/2020 – 12/31/2020
-
Corporate Publisher:
-
Abstract:Location data about U.S. heliports is often inaccurate or nonexistent in the FAA's databases, which leaves pilots and air ambulance operators with erroneous information about where to find safe landing zones. In the 2018 FAA Reauthorization Act, Congress required the FAA to collect better information from the helicopter industry under part 157, which covers the construction, alteration, activation, and deactivation of airports and heliports. At the same time, there is no requirement to report private helipads to the FAA when constructed or removed, and some public heliports do not have up-to-date records. This project developed an autonomous system that can authenticate the coordinates present in the FAA master landing site database. Our system can search for helipads in designated large areas around the country. The proposed approach is based on a convolutional neural network model that learns optimal helipad features from the data. We used the FAA's 5010 database and others to construct a benchmark database of rotorcraft landing sites. The database consists of 9,324 aerial images containing helipads, helistops, helidecks, and helicopter runways in rural and urban areas. The database also includes negative examples, i.e., satellite images, e.g., rooftop of buildings and fields that are not designated landing sites. The dataset was used to train various state-of-the-art convolutional neural network models (CNN). The outperforming model, EfficientNet-b0, achieved nearly 95% validation accuracy.
-
Format:
-
Funding:
-
Collection(s):
-
Main Document Checksum:
-
Download URL:
-
File Type:
