Sarah Boelter

Sarah Boelter

Experienced software engineer and Ph.D Candidate in Computer Science with focus in AI and Robotics. Research interests in drilling autonomy, field robotics, and arctic centered research.

Sitemap

A list of all the posts and pages found on the site. For you robots out there, there is an XML version available for digesting as well.

Pages

Posts

Future Blog Post

less than 1 minute read

Published:

This post will show up by default. To disable scheduling of future posts, edit config.yml and set future: false.

Blog Post number 4

less than 1 minute read

Published:

This is a sample blog post. Lorem ipsum I can’t remember the rest of lorem ipsum and don’t have an internet connection right now. Testing testing testing this blog post. Blog posts are cool.

Blog Post number 3

less than 1 minute read

Published:

This is a sample blog post. Lorem ipsum I can’t remember the rest of lorem ipsum and don’t have an internet connection right now. Testing testing testing this blog post. Blog posts are cool.

Blog Post number 2

less than 1 minute read

Published:

This is a sample blog post. Lorem ipsum I can’t remember the rest of lorem ipsum and don’t have an internet connection right now. Testing testing testing this blog post. Blog posts are cool.

Blog Post number 1

less than 1 minute read

Published:

This is a sample blog post. Lorem ipsum I can’t remember the rest of lorem ipsum and don’t have an internet connection right now. Testing testing testing this blog post. Blog posts are cool.

portfolio

publications

Understanding Drill Data for Autonomous Application

Published in 23rd International Conference on Autonomous Agents and Multi-Agent Systems MASSpace Workshop, 2024

In high-risk, high-cost environments like Mars it is necessary for robotic agents to either respond to or reason through potential indicators of trouble before they escalate to problems that effectively mean mission failure. Currently, no broadly applicable solution exists to give a complicated and specialized agent like The Regolith and Ice Drill for Exploring New Terrain (TRIDENT) the ability to understand the rate of its progress on a task or the situational awareness to know when a situation might escalate to a drilling fault. We examined logged data from previous field experiences to better understand potential drilling faults TRIDENT will have to reason through. We applied time series analysis techniques to determine what trends in the data exist during faults and if change point analysis or other machine learning methods could be used to predict faults of the drill.

Recommended citation: S. Boelter, E. Temesgen, B. Glass, M. Gini (2024). "Understanding Drill Data for Autonomous Application." 23rd International Conference on Autonomous Agents and Multi-Agent Systems MASSpace Workshop.
Download Paper

Automation and AI Technology Development for Planetary Drilling

Published in 2024 Committee on Space Research Scientific Assembly, 2024

Future planetary surface sampling missions, such as delving past the near-surface ice layers on Mars in search of organics and possibly signs of past/extant life, will require lightweight, low-mass planetary drilling and sample handling. Unlike terrestrial drills, these exploration drills must work dry (without drilling muds or gas), blind (no prior local or regional seismic or other surveys), and light (very low downward force or weight on bit, and perhaps 100 W available from solar power or batteries). Given the lightspeed transmission delays to Mars and outward, an exploratory planetary drill cannot be controlled directly from Earth. Drills that penetrate deeper than a few centimeters are likely to get stuck if operated open-loop (the MSL drill only penetrates 5 cm, and the MER Rock Abrasion Tools 5 mm by comparison), so some form of local drill control is required. In the relatively near-term, human crews cannot be presumed to be available for surface instrument teleoperation. Therefore highly automated drill and sample-transfer operations will be required, to explore the subsurface with the ability to safe robotic drilling systems and recover and continue on from the most probable fault conditions. Current automation, scheduling and diagnostic approaches will be discussed that roughly track the actions and roles of humans in terrestrial manual drilling operations.

Recommended citation: B. Glass, T. Stucky, T. Stevenson, S. Boelter, D. Bergman, C. Stoker, I. King (2024). "Automation and AI Technology Development for Planetary Drilling." 2024 Committee on Space Research Scientific Assembly.
Download Paper

Hierarchical and Heterogeneous MARL for Coordinated Multi-Robot Box-Pushing Environment

Published in AAAI 2025 CMASDL Workshop, 2024

Consider an environment where at least two robotic agents are surrounded by small and large-sized objects that have to be moved from their current locations to designated goal locations. Small objects can be pushed by one robotic agent while large objects need to be pushed by two robotic agents. The objective is to minimize the total time required to move all the boxes to their respective goal locations. In this paper, we explore a solution using a cooperative Partially Observable Markov Decision Process (POMDP) utilizing Proximal Policy Optimization algorithms in a simulated environment.

Recommended citation: E. Temesgen, S. Boelter, M. Gini (2024). "Hierarchical and Heterogeneous MARL for Coordinated Multi-Robot Box-Pushing Environment." AAAI 2025 CMASDL Workshop.
Download Paper

A Concept for Planetary Drilling Autonomy

Published in ICRA Workshop on Field Robotics, 2025

In planetary environments, robotic agents must reason through faults before they escalate to mission-critical failures. No broadly applicable solution exists to give a spe- cialized agent like The Regolith and Ice Drill for Exploring New Terrain (TRIDENT) awareness for when a situation may escalate to a drilling fault. We propose building on our previous work with online time-series subspace analysis methods and percussive beat frequency detection techniques to define a Markov Decision Process to autonomously avoid drilling faults.

Recommended citation: S. Boelter, E. Temesgen, G. Brown, M. Jerez, E. Forberger, B. Glass, M. Gini (2025). "A Concept for Planetary Drilling Autonomy." ICRA Workshop on Field Robotics.
Download Paper

Geofenced Unmanned Aerial Robotic Defender for Deer Detection and Deterrence (GUARD)

Published in ICRA Workshop on Novel Approaches for Precision Agriculture and Forestry with Autonomous Robots, 2025

Wildlife-induced crop damage, particularly from deer, threatens agricultural productivity. Traditional deterrence methods often fall short in scalability, responsiveness, and adaptability to diverse farmland environments. This paper presents an integrated unmanned aerial vehicle (UAV) system designed for autonomous wildlife deterrence, developed as part of the Farm Robotics Challenge. Our system combines a YOLO-based real-time computer vision module for deer detection, an energy-efficient coverage path planning algorithm for efficient field monitoring, and an autonomous charging station for continuous operation of the UAV. In collaboration with a local Minnesota farmer, the system is tailored to address practical constraints such as terrain, infrastructure limitations, and animal behavior. The solution is evaluated through a combination of simulation and field testing, demonstrating robust detection accuracy, efficient coverage, and extended operational time. The results highlight the feasibility and effectiveness of drone-based wildlife deterrence in precision agriculture, offering a scalable framework for future deployment and extension.

Recommended citation: E. Temesgen, M. Jerez, G. Brown, G. Wilson, S. Divakarla, S. Boelter, O. Nelson, R. McPherson, M. Gini (2025). "Geofenced Unmanned Aerial Robotic Defender for Deer Detection and Deterrence (GUARD)." ICRA Workshop on Novel Approaches for Precision Agriculture and Forestry with Autonomous Robots.
Download Paper

Planetary Analog Drill Automation Tests at Haughton Crater

Published in Lunar Surface Innovation Consortium (LSIC) Spring Meeting, 2025

Recommended citation: B. Glass, S. Boelter, V. Vendiola, C. Fortuin, T. Stucky, C. Stoker (2025). "Planetary Analog Drill Automation Tests at Haughton Crater." Lunar Surface Innovation Consortium (LSIC) Spring Meeting.
Download Paper

Fast Vibration-Based Autonomous Fault Detection for Extraterrestrial Drills

Published in Resource Constrained Robotics Workshop at 21st Robotics: Science and Systems (RSS) Conference, 2025

Drilling is essential to characterize the subsurface of extraterrestrial bodies and to gather samples that enable further understanding and exploration of our solar system. However, extraterrestrial drills are prone to structural failure due to the extremities in environments beyond earth, and require robust autonomous operation to prevent structural damage and achieve mission objectives. This is challenging to accomplish with conventional autonomy methods in a low resource planetary environment. While previous work utilizing structural health management (SHM) techniques excelled at detecting most on- coming faults, it required creation and validation of complex dynamic models, operation of seven neural nets in parallel, nor did it address the percussive behavior seen in modern lunar drills or provide diagnostics quickly enough to catch faults that onset in under 20 seconds. Rather than seeking to create a new resource intensive and complex learner, we seek to create lightweight models to run alongside existing systems. This work enables the use of previous SHM research by providing an unsupervised ensemble-learning method to identify percussive beats in drill vibration applicable to computation and energy constrained space environments, and shows that the frequencies present during percussive strikes can be used to perform high-frequency health diagnostics.

Recommended citation: E. Forberger, S. Boelter, B. Glass, M. Gini (2025). "Vibration-Based Autonomous Fault Detection for Extraterrestrial Drills." Resource Constrained Robotics Workshop at 21st Robotics: Science and Systems (RSS) Conference.
Download Paper

Fault Prediction in Drilling using Subspace Analysis Techniques

Published in Intelligent Autonomous Systems 19, Proc. 19th International Conf IAS-19, 2025

In remote planetary environments, robotic agents must re- spond to or reason through faults before they escalate to mission-critical failures. No broadly applicable solution exists to give a specialized agent like The Regolith and Ice Drill for Exploring New Terrain (TRIDENT) situational awareness for when a situation may escalate to a drilling fault. We propose a new online time-series subspace analysis method, Entan- gled Singular Spectrum Transformation (ESST), to better predict and analyze faults using online data produced by the TRIDENT drill. We evaluate performance against other online subspace analysis techniques to determine the optimal detection method for sudden drilling faults.

Recommended citation: S. Boelter and L. Weber, B. Glass, M. Gini. (2025). "Fault Prediction in Drilling using Subspace Analysis Techniques." Intelligent Autonomous Systems 19, Proc. 19th International Conf IAS-19.
Download Paper

Diagnostics for Drilling Fault Prediction in Planetary Drills

Published in 9th International Symposium on Ice Drilling Technology, 2025

Recommended citation: S. Boelter, G. Brown, T. Stucky, E. Forberger, G. Wilson, M. Gini, B. Glass. (2025). "Diagnostics for Drilling Fault Prediction in Planetary Drills." 9th International Symposium on Ice Drilling Technology.

Planetary Drill Automation Tests at Analog Sites

Published in AGU Annual Meeting, 2025

Recommended citation: B. Glass, S. Boelter, C. Fortuin, T. Stucky, C. Stoker. (2025). Planetary Drill Automation Tests at Analog Sites. In AGU Annual Meeting. December 2025, New Orleans, USA.

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

Published in IEEE International Conference on Robotics and Automation (ICRA), 2026

In extraterrestrial planetary environments, computing, energy, and environmental constraints require robotic agents to complete tasks without supervision. For specialized extraterrestrial 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 estimate drilling faults using drill avionics telemetry. This work introduces 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 environment with composite materials and then in a Mars planetary analog site in the Canadian High Arctic.

Recommended citation: S. Boelter, G. Brown, E. Temesgen, L. Weber, T. Stucky, B. Glass, M. Gini. (2026). Model-Free Subsurface Anomaly Detection using Subspace Analysis Techniques for Sparse Telemetry for Extraterrestrial Drilling Robots. Proceedings IEEE International Conference on Robotics and Automation (ICRA). June 2026. Accepted for Publication
Download Paper

talks

teaching

Teaching experience 1

Undergraduate course, University 1, Department, 2014

This is a description of a teaching experience. You can use markdown like any other post.

Teaching experience 2

Workshop, University 1, Department, 2015

This is a description of a teaching experience. You can use markdown like any other post.