We present the first publicly-available RGB-thermal dataset designed for aerial robotics operating in natural environments. Our dataset captures a variety of terrain across the United States, including rivers, lakes, coastlines, deserts, and forests, and consists of synchronized RGB, thermal, global positioning, and inertial data. We provide semantic segmentation annotations for 10 classes commonly encountered in natural settings in order to drive the development of perception algorithms robust to adverse weather and nighttime conditions. Using this dataset, we propose new and challenging benchmarks for thermal and RGB-thermal (RGB-T) semantic segmentation, RGB-T image translation, and motion tracking. We present extensive results using state-of-the-art methods and highlight the challenges posed by temporal and geographical domain shifts in our data. The dataset and accompanying code is available at https://github.com/aerorobotics/caltech-aerial-rgbt-dataset.
",
"doi": "10.1007/978-3-031-73036-8_14",
"isbn": "978-3-031-73035-1",
"publisher": "Springer",
"place_of_publication": "Cham",
"publication": "Computer Vision \u2013 ECCV 2024",
"publication_date": "2024-11-21",
"pages": "236-256"
},
{
"id": "authors:83c93-zb655",
"collection": "authors",
"collection_id": "83c93-zb655",
"cite_using_url": "https://authors.library.caltech.edu/records/83c93-zb655",
"type": "article",
"title": "Learning-Based Minimally-Sensed Fault-Tolerant Adaptive Flight Control",
"author": [
{
"family_name": "O'Connell",
"given_name": "Michael",
"orcid": "0000-0001-6681-8823",
"clpid": "O'Connell-Michael-T"
},
{
"family_name": "Cho",
"given_name": "Joshua",
"orcid": "0009-0000-5482-0010",
"clpid": "Cho-Joshua"
},
{
"family_name": "Anderson",
"given_name": "Matthew",
"orcid": "0000-0001-8884-3448",
"clpid": "Anderson-Matthew-J"
},
{
"family_name": "Chung",
"given_name": "Soon-Jo",
"orcid": "0000-0002-6657-3907",
"clpid": "Chung-Soon-Jo"
}
],
"abstract": "\n
\n
Many multirotor aircraft use redundant configurations to maintain control in the event of an actuator failure. Due to the redundancy of the system, fault isolation is inherently difficult and further compounded by complex interacting aerodynamics of the propellers, wings, and body. This letter presents a novel sparse failure identification and control correction method that does not require direct fault sensing, and instead utilizes only the vehicle's dynamic response. The method couples an \u2113\u2081-regularized representation of the failure with a deep neural network to effectively isolate faults and improve tracking control in highly dynamic environments with unmodeled aerodynamic effects and unknown actuator failures. The method also includes a control re-allocation scheme which corrects for the identified faults while maximizing control authority and maintaining nominal performance characteristics. Experimental results demonstrate the method's ability to maintain control of a multirotor aircraft by isolating motor failures and reallocating control, improving position tracking by 48 % over the baseline. This letter contributes to the development of robust fault detection and control strategies for over-actuated aircraft.
\n
\n
\n
\n
To establish a long-term human presence on the Moon, NASA will rely on the in-situ resource utilization, or ISRU, of native ores in its lunar metal production pipeline. Therefore, developing a practical, sustainable, and affordable method to extract useful metals from these native ores is essential for all future lunar manufacturing and construction efforts. This paper presents the Solar Powered Apparatus for Melting Materials on the Moon, or SPAMMM, as a metal extraction solution to be implemented on the Moon’s South Pole, one of the targeted Artemis landing sites. SPAMMM consists of a central “power tower” that contains gas injection valves, a depressurization valve, a distillation chamber, a precipitate chamber, and a furnace that separates metals from their initial ores via concentrated solar radiation from a foldable parabolic mirror. In order to leverage the uniquely shallow angle of incident light at the South Pole, this parabolic mirror is rotated about the furnace to track the weeks-long daylight. To maximize efficiency, the radius of the mirror’s path has been designed to be the mirror’s focal length. The paper is divided into the following sections: (1) Problem Statement, where the need for ISRU in the lunar metal production pipeline is explained; (2) Project Description, where the requirements, concept of operations, and design overview are detailed; (3) Ore Extraction, which discusses the power tower design and chemical processes behind extracting metal from specific lunar ores; (4) Rotating Parabolic Concentrator, which details the design, optical specifications, and thermal capabilities of the rotating parabolic mirror; (5) Verification Testing Program, which outlines the system and subsystem testing plan that will bring SPAMMM to a Technology Readiness Level of 5 and details any additional consideration for SPAMMM’s lunar implementation. If awarded funding through the 2023 NASA BIG Idea Challenge, future work will involve developing, designing, and validating every subsystem of SPAMMM in the relevant environment.
\n
\n