The density of satellites in Earth’s orbit has surged exponentially in recent years, driven by the reduced costs of small satellites. This has enabled governments, researchers, and private companies to launch and operate approximately 2,877 satellites in orbit in 2023 alone. This includes increased activity of geostationary satellites in Earth’s orbit (GEO), which bring globally impactful technologies such as high-speed internet and climate monitoring. However, alongside the numerous benefits of these satellite technologies come heightened safety and security risks, as well as environmental concerns. More precise and efficient methods for monitoring and modeling satellite behavior are urgently needed to prevent collisions and other disasters.
To address this challenge, MIT’s Astrodynamics, Space Robotics, and Control Laboratory (ARCLab) has launched the MIT ARCLab Prize for AI Innovation in Space: a unique competition that calls on participants to leverage AI to characterize satellite Patterns of Life (PoL) — the long-term behavioral narrative of a satellite in orbit — using purely passively collected information. Following last fall’s call for participants, 126 teams utilized machine learning to create algorithms that label and timestamp the behavior modes of GEO satellites over a six-month period, competing for accuracy and efficiency.
Supported by the U.S. Air Force-MIT AI Accelerator, the challenge offers a total of $25,000 in prizes. A panel of judges from ARCLab and the MIT Lincoln Laboratory evaluated the submissions based on clarity, novelty, technical depth, and reproducibility, scoring each entry out of 100 points. The judges have now announced the winners and finalists:
First Prize: David Baldsiefen — Team Hawaii2024
With a winning score of 96, Baldsiefen will receive $10,000 and an invitation to join the ARCLab team for a poster presentation at the Advanced Maui Optical and Space Surveillance Technologies (AMOS) conference in Hawaii this fall. One evaluator noted, « Clear and concise report, with very good ideas such as locator label encoding. Decisions on architectures and feature engineering are well-motivated. The provided code is also well-documented and structured, allowing for easy reproducibility of the experimentation. »
Second Prize: Binh Tran, Christopher Yeung, Kurtis Johnson, Nathan Metzger — Team Millennial-IUP
With a score of 94.2, Millennial-IUP will receive $5,000 and also join the ARCLab team at the AMOS conference. An evaluator stated, « The chosen models were sensible and justified, and they made impressive efforts in efficiency gains… They used physics to inform their models, and it seemed reproducible. Overall, it was a concise and easy-to-follow report, without too much jargon. »
Third Prize: Isaac Haik and François Porcher — Team QR_Is
With a score of 94, Haik and Porcher will share the third prize of $3,000 and will also be invited to the AMOS conference with the ARCLab team. An evaluator noted, « This informative and interesting report describes the combination of ML and signal processing techniques convincingly, assisted by informative plots, tables, and sequence diagrams. The author identifies and describes a modular approach to class detection and their evaluation of feature utility, correctly identifying that not all features are equally useful across classes… Any lack of mission expertise is compensated by a clear and detailed discussion of the benefits and pitfalls of the methods they used and what they learned. »
Teams ranked fourth through seventh will each receive $1,000 and a certificate of excellence.
« The goal of this competition was to foster an interdisciplinary approach to problem-solving in the space domain by inviting AI development experts to apply their skills in this new context of orbital capability. And all of our winning teams truly delivered: they brought their technical skills, innovative approaches, and expertise to a series of very impressive submissions, » says Professor Richard Linares, who leads ARCLab.
Active Modeling with Passive Data
Throughout the orbital time of a GEO satellite, operators issue commands to place it in different behavior modes: station-keeping, longitudinal maneuvers, end-of-life behaviors, etc. Satellite Patterns of Life (PoL) describe the in-orbit behavior composed of sequences of natural and unnatural behavior modes.
ARCLab has developed a groundbreaking benchmarking tool for the characterization of geosynchronous satellite life patterns and created the Satellite Pattern of Life Identification Dataset (SPLID), comprising data from real and synthetic space objects. Challenge participants used this tool to create AI algorithms to map a satellite’s in-orbit behaviors.
The goal of the MIT ARCLab Prize for AI Innovation in Space is to encourage technologists and enthusiasts to bring innovation and new skills to the well-established challenges of aerospace. The team aims to host the competition in 2025 and 2026 to explore other topics and invite AI experts to apply their skills to new challenges.
