← Back ICRA 2026

Model-Free Subsurface Anomaly Detection Using Subspace Analysis Techniques for Sparse Telemetry for Extraterrestrial Drilling Robots

Sarah Boelter, Greta Brown, Ebasa Temesgen, Lucas Weber, Thomas Stucky, Brian Glass, Maria Gini

PDF

AI summary

Key figure (auto-extracted from paper)

An unsupervised subspace analysis method successfully detects drilling faults in real-time using sparse telemetry, eliminating the need for extensive training data in extraterrestrial environments.

Subsurface anomaly detection Model-free fault detection Enhanced Singular Spectrum Transformation Planetary drilling autonomy Real-time telemetry analysis Mars analog testing

Problem

Robotic planetary drills require early fault detection to prevent hardware failure, but conventional supervised machine learning is impractical due to severe constraints on computing, energy, and the scarcity of fault training data.

Approach

The authors implement an online, model-free change point detection algorithm that continuously analyzes streaming motor telemetry to identify subsurface anomalies without prior fault examples.

Key results

Deployed an unsupervised ESST fault detection algorithm on the TRIDENT extraterrestrial drill
Achieved high detection accuracy (F1 > 0.6) for motor torque, rate of penetration, and acoustic telemetry in controlled lab tests
Validated robust real-time anomaly detection across diverse geological conditions at a Mars-analog field site
Demonstrated significant statistical separation (Cohen’s D > 0.8) between normal and faulty telemetry signals without manual calibration

Why it matters

Enables reliable, autonomous fault detection for future planetary drilling missions where communication delays and limited computational resources preclude Earth-based supervision or data-heavy AI models.

Abstract

In extraterrestrial planetary environments, com- puting, energy, and environmental constraints require robotic agents to complete tasks unsupervised. For specialized extrater- restrial robotic drilling agents there is no broadly applicable solution to detect drilling faults as they happen, before the fault escalates to hardware failure. We build upon previous work with time-series subspace analysis methods to to estimate drilling faults using drill avionics telemetry. This work intro- duces a subsurface anomaly detection method for planetary drilling robots and further evaluates the robustness of our time-series subspace analysis method. We implemented this novel fault and anomaly detection method on an extraterrestrial drilling robot and evaluated it first in a controlled lab environ- ment with composite materials and then in a Mars planetary analog site in the Canadian High Arctic.

Index terms

Field Robots Space Robotics and Automation Mining Robotics