Rotorcraft Landing Sites Identification – Scaling and Generalization of the AI Model

Download Final Report

CAIT project no.: CAIT-UTC-REG54

Fiscal Year: 2020/2021

Status: Final

Principal investigator(s): Ghulam Rasool, Ph.D. (PI), Rowan University
Nidhal Bouaynaya, Ph.D. (Co-PI), Rowan University

Performing organization(s): Rowan University

Managing organization: Rutgers CAIT

In cooperation with: William J. Hughes Technical Center, FAA
Partner project manager: Charles (Cliff) Johnson, Research Lead, Rotorcraft, Unmanned Aircraft Systems (UAS)

Supported by: USDOT-OST-R

UTC, grant, or agreement no.: 69A3551847102


The primary goal of this proposal is to address the challenging problem of automatic identification of helipads and landing sites using the machine and deep learning algorithms. This project’s deliverable is an AI-based system for the identification of helipads, heliports, and landing site infrastructure from satellite images.

The intended outcome of the AI model is to automate the process of identification of landing sites for rotorcrafts from the Google Earth satellite imagery. This system is expected to achieve landing site identification accuracy equal to or higher than that of a trained human operator at a fraction of time and resources. Once developed, the AI system would allow the FAA to regularly update its databases without delays and, as a result, the databases of FAA could be used by any mission, including “Helicopter Air Ambulance missions to rural communities.”