The First Moon Landing Wasn’t Apollo — And We Just Found It

Quick Overview

A recent study utilizing a novel machine learning algorithm (YOLO-ETA) suggests the possible re-identification of the Soviet Luna 9 landing site, which occurred in 1966, by analyzing high-resolution Lunar Reconnaissance Orbiter Camera (LROC) imagery and finding high-confidence matches for discarded spacecraft hardware, contrasting with the fact that the US Apollo missions' landing sites are already confirmed.

Key Points: A research group used a novel machine learning algorithm, YOLO-ETA (You-Only-Look-Once – Extraterrestrial Artefact), adapted from TinyYOLOv2, to search LROC imagery for anthropogenic objects on the Moon. The algorithm was trained on Apollo landing site data and achieved a balanced precision-recall (F1 = 0.60) and an 80% mean confidence score for lander detections in previously unseen images. Applying the model to a 5x5 km region surrounding the historically uncertain Luna 9 landing area yielded several high-confidence detections of artificial objects near 7.03° N, -64.33° E. The analysis indicates that the candidate site's horizon geometry is potentially consistent with Luna 9's topographical panoramas, suggesting a possible location for the first Soviet Moon landing in 1966. The researchers also successfully detected hardware remnants from the Apollo missions (like the Lunar Module, flag, LRRR, PSEP) and the Soviet Luna 16 lander, demonstrating the system's effectiveness. The study suggests this AI-driven methodology can support future orbital surveys of lunar artifacts and surface details, potentially resolving ambiguities in historical landing locations.

Context: The video discusses a recent scientific paper published in npj Space Exploration (2026) detailing the potential identification of the landing site for the Soviet Luna 9 probe, the first spacecraft to achieve a soft landing on the Moon in 1966. This contrasts with the US Apollo landing sites, which are already well-documented and imaged by NASA's Lunar Reconnaissance Orbiter (LRO). The key development is the application of a custom machine learning system called YOLO-ETA to search historical LROC images for small, artificial debris left behind by these early missions.