Evaluating Single-Camera Range Estimation for Autonomous Air-to-Air Refueling During Rendezvous Approach Phase

The U.S. Navy plans to dramatically increase the number of uncrewed aerial systems (UASs) within its carrier air wings. These UASs are expected to operate in GPS-denied and RF-denied environments, requiring fully autonomous capabilities. A key enabler is the ability to perform autonomous air-to-air refueling (A3R) without reliance on external positioning systems. This research investigates the viability of using a single vision-based camera—originally intended for drogue docking—to estimate range during the rendezvous phase of A3R. A deep neural network (DNN)–based monocular ranging system was adapted and evaluated in the University of Maryland’s (UMD’s) Omni-domain Autonomous Systems Integration Space (OASIS), a 4,800 square foot motion-capture environment. A YOLO11m DNN was trained to identify the aerial refueling coupler and drogue basket at ranges from 5–50 ft. A similar-triangles ranging algorithm produced real-time position estimates that were compared against truth data. Across 3,000 test points, the system achieved near-zero mean error for coupler ranging from 5–17 ft, and maintained errors under 10 in. up to 34 ft. When transitioning from coupler to drogue basket object detection, the mean error remained below 10 in. up to 45 ft. Pixel-level algorithm sensitivity increases estimation error nonlinearly with range, indicating the need to transition to detecting larger objects. These results demonstrate that a single monocular optical payload can support both docking and a substantial portion of the rendezvous phase, provided that the detected object scales appropriately with range. Future work will investigate DNN detection of larger tanker features to extend the capability to longer ranges.