What is it about?
The population of low Earth orbit (LEO) has skyrocketed over the last few years because of anti-satellite tests, in-space collisions, and the proliferation of micro-satellite constellations. The systems that are used to keep track of—and, therefore, avoid collisions between—these objects are struggling to keep up with this growth. To help reduce the strain on these systems, we've developed a novel optimization formulation to improve the way we task and schedule the sensors used to monitor resident space objects in LEO.
Featured Image
Photo by Yasuaki Uechi on Unsplash
Why is it important?
Recent efforts to improve sensor coordination for tracking objects in space have focused on geosynchronous Earth orbit, which has orders of magnitude fewer objects than LEO, or used approximate methods, which are inherently suboptimal. Our approach is based on the vehicle routing problem, a classical optimization technique that can provide guarantees of optimality. To enable efficient optimization at LEO operational scales, we pre-process the nonlinear dynamics and constraints of the problem into matrices of 1s and 0s, which are leveraged by our integer program formulation. We show that our approach, given 4 hours on a high-performance computing cluster, can find global and regional optima in scenarios with 27 ground-based sensors tasked to track 10,000 targets over a 24-hour time horizon.
Perspectives
A lot has happened in low Earth orbit since the SP Tasker algorithm went online in 2005. With prominent in-space collisions, anti-satellite tests, and microsatellite constellations taking to the skies, the population of LEO has at least doubled—if not tripled—over the last twenty years. We're long overdue for a new approach to sensor orchestration, and this article offers a scalable approach toward addressing that challenge.
Allan Shtofenmakher
Massachusetts Institute of Technology
Read the Original
This page is a summary of: Vehicle Routing Problem Formulation for Efficient Tracking of Objects in Low Earth Orbit, January 2025, American Institute of Aeronautics and Astronautics (AIAA),
DOI: 10.2514/6.2025-0533.
You can read the full text:
Resources
Performance Demonstration with 10,000 Objects
This video demonstrates the performance of our approach in a scenario with 10,000 objects tracked by 27 sensors over 24 hours. Both the rolling time horizon and the global optimization approaches are effective, though the latter offers marginally better performance.
Optimal Tasking and Scheduling of Satellite Constellations for Space Situational Awareness
Related conference paper on "Optimal Tasking and Scheduling of Satellite Constellations for Space Situational Awareness," presented by A. Shtofenmakher at the AMOS Conference in Maui, Hawai'i, on Wednesday, September 17, 2025, at 6:00 PM HST.
Contributors
The following have contributed to this page
