Global Task-aware Fault Detection, Identification For On-Orbit Multi-Spacecraft Collaborative Inspection

TL;DR

This work tackles fault detection and isolation in a multi-spacecraft ensemble performing a collaborative inspection by tying network-level task objectives to local agent behavior through a global cost functional $\mathcal{H}$. It introduces a global-to-local FDI architecture, decomposes $\mathcal{H}$ into per-agent contributions with a time-varying consensus term, and employs a gradient-based fault metric using higher-order derivatives to detect and classify both global and agent-level faults. An adaptive, penalty-based threshold $\tau_i(t)$ is designed to account for time-varying task dynamics, with actuator and sensor faults identified in simulation of a low-Earth orbit inspection mission. The approach supports graceful degradation and autonomous recovery by informing local actions via the central FDI system, demonstrating robustness to time-varying graphs and intermittent communication. $\mathcal{H}$ is central to the methodology, with $\mathcal{H}_{\mathrm{POI}}(\mathbf{s}) = \left( w^{-1} + \sum_{\mathbf{p}\in\mathcal{P}} f(\mathbf{p},\mathbf{s})^{-1} \right)^{-1}$ and $\mathcal{H} = \sum_{\mathbf{s}\in\text{POIs}} \mathcal{H}_{\mathrm{POI}}(\mathbf{s}) \phi(\mathbf{s})$, enabling per-agent diagnostics through $\mathcal{H}_i(t)$ and its predicted counterpart $\mathcal{H}_i^{\text{pred}}(t)$.

Abstract

In this paper, we present a global-to-local task-aware fault detection and identification algorithm to detect failures in a multi-spacecraft system performing a collaborative inspection (referred to as global) task. The inspection task is encoded as a cost functional $\costH$ that informs global (task allocation and assignment) and local (agent-level) decision-making. The metric $\costH$ is a function of the inspection sensor model, and the agent full-pose. We use the cost functional $\costH$ to design a metric that compares the expected and actual performance to detect the faulty agent using a threshold. We use higher-order cost gradients $\costH$ to derive a new metric to identify the type of fault, including task-specific sensor fault, an agent-level actuator, and sensor faults. Furthermore, we propose an approach to design adaptive thresholds for each fault mentioned above to incorporate the time dependence of the inspection task. We demonstrate the efficacy of the proposed method empirically, by simulating and detecting faults (such as inspection sensor faults, actuators, and sensor faults) in a low-Earth orbit collaborative spacecraft inspection task using the metrics and the threshold designed using the global task cost $\costH$.

Figures (7)

• Figure 1: Global task-aware fault detection and isolation for a distributed spacecraft netwrok.
• Figure 2: An overviwe of the type of global and local faults detected and identified using the proposed FDI architecture, metrics and the threshold.
• Figure 3: real-time vs. Expected cost under actuator attack I (left); Behavior of fault signal (right).
• Figure 4: Real-time vs. Expected cost under actuator attack II (left); Behavior of fault signal (right).
• Figure 5: Visible set of POIs for a spacecraft (left); $\epsilon-$neighborhood constructed around POI with maximum variance (right).