Master Thesis Proposal

Lucien Lamarre Saubusse

(Faculty Advisor: Professor Dimitri Mavris)

"Optimal Fleet Design and Operations for Collection-as-a-Service (CAAS) Orbital Satellite Recovery in Low Earth Orbit"

Tuesday, July 28

9:00 - 11:00 a.m.

Weber, CoVE

Abstract:

Low Earth Orbit has become an increasingly congested and commercially active region, and the accumulation of satellites reaching their end of life raises concerns for both the sustainability and the economics of future operations. Active debris removal and on-orbit servicing have been studied for several decades as possible responses, and a substantial body of work now exists on the underlying technologies and mission concepts. Much of this literature, however, treats removal as a technical or regulatory problem rather than as a logistics enterprise that must be profitable to be realized at scale. Collection-as-a-Service (CAAS), in which orbital platforms coordinate contracted tugs to recover end-of-life satellites and return them to Earth, has been proposed as one such commercial concept. This suggests the need for a design capability that treats CAAS not as a single mission but as a fleet to be sized, placed, and operated under economic objectives.

System design research has long shown that most lifecycle cost is committed in the early design stages, where configuration and preliminary operations are fixed with limited information. Orbital logistics studies have addressed related problems, including facility location, vehicle routing, and payload selection, but these are often examined separately or under the more forgiving geometry of geostationary orbit. In LEO, the design space is considerably harder to navigate: orbital plane changes are prohibitively expensive in propellant, and the oblateness of the Earth induces a differential nodal drift that continuously separates orbital planes over time. These effects couple platform placement, tug routing, and recovery scheduling in ways that resist the sequential treatment common in prior work. This motivates a framework in which orbital mechanics and commercial operations are considered concurrently rather than in isolation.

This proposal presents the questions and candidate methods through which such a framework may be built. The first investigation concerns the overall architecture, and posits that the coupled placement, routing, and selection problem can be decomposed into an outer genetic algorithm that proposes platform orbits and an inner mixed-integer linear program that returns the operating profit. The second investigation concerns the treatment of orbital dynamics, and proposes that the nonlinear transfer costs can be precomputed into a fixed, time-expanded network so that the inner optimization remains fully linear. The third investigation concerns fidelity, and asks under which conditions the coplanar and instantaneous-transfer assumptions remain acceptable, a question that can be tested against higher-fidelity propagation of the same scenarios.

Using the results of these investigations, the improvement offered to CAAS design methodology can be demonstrated on representative test cases. A first, simplified scenario is used to show how the profit and recovery metrics produced by the framework can inform early configuration selection. A second scenario, closer to a realistic constellation-disposal setting, is used to exercise the joint sizing, placement, and routing of a multi-platform fleet. Both cases aim to show that a computationally tractable and reasonably robust design approach can be extended to commercial orbital logistics in LEO.

Committee:
Dr. Dimitri Mavris (advisor), School of Aerospace Engineering
Dr. Tristan Sarton du Jonchay, School of Aerospace Engineering
Dr. Olivia Fischer, School of Aerospace Engineering