Ravi teja Nallapu, University of Arizona/ SpaceTREx; Himangshu Kalita, Univ. of Arizona/SpaceTREx; Jekan Thangavelautham, Univ. of Arizona/SpaceTREx
Keywords: swarm constellations; meteor impact; CubeSats; Small Satellites; Decentralized Control
Abstract:
Meteor impact events such as Chelyabinsk are known to cause catastrophic impacts every few hundred years. The modern-day fallout from such event can cause large loss of life and property. There is a real need to have advanced warning of such meteor events including distinguishing them from human activity such as ballistic missile launches.
One satellite in Low Earth Orbit (LEO) may get glimpse of a meteor trail. However, a swarm constellation has the potential of performing persistent, real time global tracking of an incoming meteor. Developing a swarm constellation using a fleet of large satellites is an expensive endeavor. A cost-effective alternative is to use CubeSats and small-satellites that contain the latest, onboard computers, sensors and communication devices for low-mass and low-volume. Utilizing a swarm constellation that is launched using multiple rocket launches reduces the risk of a single rocket failure. Furthermore, additional satellites maybe warehoused in space to quickly respond to unexpected events and losses.
Our inspiration for a swarm constellation come from eusocial insects that are composed of simple individuals, that are decentralized, that operate autonomously using only local sensing and are robust to individual losses. A swarm constellation is immune to single-point failures unlike a monolithic large spacecraft and is expected to show graceful degradation in performance due to loss of nodes. In addition, one or more nodes maybe repurposed to dynamically perform task-decomposition and monitor one or more events concurrently. More nodes maybe readily added to the swarm, exploiting the latest advances in sensor and computation making the system extensible and steadily upgradeable.
In this presentation, we first analyze the capabilities of a single CubeSat node in the proposed swarm constellation and then extend that capability to a swarm. We use physics simulations to analyze the design, performance and feasibility of positioning the satellites with and without propulsion. These physical simulations are directed using automated design and control approaches using a form of Evolutionary Algorithms. Through these simulations, we analyze and determine the optimal number of nodes required to perform area coverage of North America and its surrounding. Furthermore, we analyze the performance of the swarm based on loss of one or more satellites. Finally, we analyze the impact and effectiveness of the swarm with several CubeSat/small-sat compatible propulsion technologies suited for LEO.
Our simulation work shows that by having swarms of CubeSats monitor these events from multiple locations it is possible to triangulate the position of the meteor in real-time, obtain readings on their size and possibly composition. This is further enhanced using ground observational assets. These simulations point towards a pathway for further development of this technology towards an eventual on-orbit demonstration.
Date of Conference: September 11-14, 2018
Track: Poster