About

Timothy Bretl is a Professor of Aerospace Engineering at the University of Illinois at Urbana-Champaign, where he has held faculty positions since 2006 and became a Professor in 2022. He also serves as the Severns Faculty Scholar and Associate Head of Aerospace Engineering. Bretl's educational background includes a Ph.D. in Aeronautics and Astronautics from Stanford University, earned in 2005, with a thesis focused on multi-step motion planning applied to free-climbing robots. He also holds a Master's degree in Aeronautics and Astronautics from Stanford, and both a B.A. in Mathematics and a B.S. in Engineering from Swarthmore College. His research interests encompass engineering education, rehabilitation robotics (including upper- and lower-limb prosthetic devices), brain-machine interfaces, optimal control, robotic manipulation, motion planning, and the theoretical and algorithmic foundations of robotics and automation. Bretl has contributed to the field through various publications, including chapters in books and articles in peer-reviewed journals, focusing on topics such as multi-agent control, motion planning for humanoid and lunar robots, and free-climbing robots. He has also been involved in multiple research centers and partnerships, and has served as a reviewer for research funding and academic journals.

Research topics

• Computer Science
• Artificial Intelligence
• Computer vision
• Machine Learning
• Mechanical engineering
• Physics
• Algorithm
• Geometry
• Medicine
• Mathematics education
• Classical mechanics
• Medical education
• Mathematics
• Engineering
• Pedagogy
• Psychology

Selected publications

• Estimating Force Interactions of Deformable Linear Objects from their Shapes

  arXiv (Cornell University) · 2026-02-01

  articleOpen access

  This work introduces an analytical approach for detecting and estimating external forces acting on deformable linear objects (DLOs) using only their observed shapes. In many robot-wire interaction tasks, contact occurs not at the end-effector but at other points along the robot's body. Such scenarios arise when robots manipulate wires indirectly (e.g., by nudging) or when wires act as passive obstacles in the environment. Accurately identifying these interactions is crucial for safe and efficient trajectory planning, helping to prevent wire damage, avoid restricted robot motions, and mitigate potential hazards. Existing approaches often rely on expensive external force-torque sensor or that contacts occur at the end-effector for accurate force estimation. Using wire shape information acquired from a depth camera and under the assumption that the wire is in or near its static equilibrium, our method estimates both the location and magnitude of external forces without additional prior knowledge. This is achieved by exploiting derived consistency conditions and solving a system of linear equations based on force-torque balance along the wire. The approach was validated through simulation, where it achieved high accuracy, and through real-world experiments, where accurate estimation was demonstrated in selected interaction scenarios.

  PublisherOA PDF

• Estimating Force Interactions of Deformable Linear Objects from their Shapes

  Open MIND · 2026-02-01

  preprint

  This work introduces an analytical approach for detecting and estimating external forces acting on deformable linear objects (DLOs) using only their observed shapes. In many robot-wire interaction tasks, contact occurs not at the end-effector but at other points along the robot's body. Such scenarios arise when robots manipulate wires indirectly (e.g., by nudging) or when wires act as passive obstacles in the environment. Accurately identifying these interactions is crucial for safe and efficient trajectory planning, helping to prevent wire damage, avoid restricted robot motions, and mitigate potential hazards. Existing approaches often rely on expensive external force-torque sensor or that contacts occur at the end-effector for accurate force estimation. Using wire shape information acquired from a depth camera and under the assumption that the wire is in or near its static equilibrium, our method estimates both the location and magnitude of external forces without additional prior knowledge. This is achieved by exploiting derived consistency conditions and solving a system of linear equations based on force-torque balance along the wire. The approach was validated through simulation, where it achieved high accuracy, and through real-world experiments, where accurate estimation was demonstrated in selected interaction scenarios.

  DOI

• Data for The MagPIE2 Dataset: Magnetic Field-Based Mapping, Localization, and SLAM

  Open MIND · 2026-01-01

  datasetSenior author

  If you use this dataset, please cite both the dataset and the associated data paper (bibtex is below). @ARTICLE{11386847, author={Hanley, David and Lee, Jongwon and Choi, Su Yeon and Bretl, Timothy}, journal={IEEE Transactions on Instrumentation and Measurement}, title={The MagPIE2 Dataset for Mapping, Localization, and Simultaneous Localization and Mapping Using Magnetic Fields}, year={2026}, volume={}, number={}, pages={1-1}, keywords={Magnetometers;Magnetic field measurement;Magnetic fields;Pedestrians;Location awareness;Buildings;Simultaneous localization and mapping;Measurement errors;Hardware;Calibration;Localization;mapping;SLAM;dataset;benchmark;magnetometer;magnetic field}, doi={10.1109/TIM.2026.3662919}} We present a dataset for the evaluation of magnetic field-based robotic and pedestrian localization, mapping, and SLAM methods. This dataset contains magnetometer and inertial measurement unit data collected from inside three buildings both a pedestrian and a ground robot. Data were collected at different heights simultaneously, both with and without changes in the placement of objects that may affect magnetometer measurements. In total, approximately 689 square meters of floor space was covered by this dataset. This dataset is archivally stored. We provide a GitHub site which is meant to serve as a forum to post issues with the dataset, share code using the dataset, and to resolve problems: https://github.com/hanley6/MagPIE2Forum Note that while the dataset is meant to be permanently stored, this forum is not meant to guarantee perennial support and its existence will be dependent on the policies of GitHub. <b>How is the dataset organized?</b> The data is divided into the following parts at a high level and more detailed information can be found in the Readme: 1. The walking portion of the dataset: CSL_WLK.zip, DCL_WLK.zip, Talbot_WLK.zip, and WLK_Misc.zip. 2. The robot portion of the dataset: Robot_Dataset.zip. 3. Motor interference tests: Motor_Interference_Test.zip. 4. Ground truth evaluation: Ground_Truth_Evaluation.zip. 5. Quick start results: Quick_Start_Results.zip. <b>How is data recorded and stored?</b> Data is generally collected in the form of ROS bag files. Each ROS bag has Intel Realsense camera images, magnetometer readings, IMU readings, timestamps, and more as applicable for each file in the dataset. Each bag file has an associated metadata file written as a YAML file. This contains general information about each bag file including the start and stop time, who collected the bag file (during the pedestrian portion of the dataset), and the approximate location where data was collected. In several cases, additional comma separated (csv) files of the dataset where included either as a convenient supplement to ROS bag files (e.g., csv files of magnetometer calibration data) or because they serve as human readable quick start results. <b>How does one set up and run files on the dataset?</b> The files are stored in ROS bags and are, therefore, meant to be run using the Robot Operating System. Information regarding how to use the Robot Operating System as well as installation instructions are available at: https://ros.org/

  DOI

• The MagPIE2 Dataset for Mapping, Localization, and Simultaneous Localization and Mapping Using Magnetic Fields

  IEEE Transactions on Instrumentation and Measurement · 2026-01-01

  articleSenior author

  This article introduces the Magnetic Positioning Indoor Estimation 2.0 (MagPIE2) dataset, designed to evaluate localization, mapping, and simultaneous localization and mapping (SLAM) methods that rely on the use of magnetic fields. Prior work has shown that contemporary localization methods that use magnetic fields are not robust to height changes that may commonly occur during use by pedestrians and robots. Our dataset contains measurements from magnetometers and inertial measurement units collected inside multiple buildings at multiple heights with both a pedestrian and a ground robot. We used several sensors for ground truth, including visual-inertial SLAM, red-green-blue (RGB) and depth images. We also provide sensor calibration measurements and suggested calibration values. Unlike some previous datasets, MagPIE2 includes trials where ferromagnetic objects (like tables, chairs, and computers) have been moved. This dataset is available for download at https://doi.org/10.13012/B2IDB-0228701_V1.

  PublisherDOI

• Accurate Simulation and Parameter Identification of Deformable Linear Objects using Discrete Elastic Rods in Generalized Coordinates

  2025-10-19

  article

  This paper presents a fast and accurate model of a deformable linear object (DLO) – e.g., a rope, wire, or cable – integrated into an established robot physics simulator, MuJoCo. Most accurate DLO models with low computational times exist in standalone numerical simulators, which are unable or require tedious work to handle external objects. Based on an existing state-of-the-art DLO model – Discrete Elastic Rods (DER) – our implementation provides an improvement in accuracy over MuJoCo’s own native cable model. To minimize computational load, our model utilizes force-lever analysis to adapt the Cartesian stiffness forces of the DER into its generalized coordinates. As a key contribution, we introduce a novel parameter identification pipeline designed for both simplicity and accuracy, which we utilize to determine the bending and twisting stiffness of three distinct DLOs. We then evaluate the performance of each model by simulating the DLOs and comparing them to their real-world counterparts and against theoretically proven validation tests.

  PublisherDOI

• The Lazy Student’s Dream: ChatGPT Passing an Engineering Course on Its Own

  IFAC-PapersOnLine · 2025-01-01

  articleOpen accessCorresponding

  This paper presents a comprehensive investigation into the capability of Large Language Models (LLMs) to successfully complete a semester-long undergraduate control systems course. Through evaluation of 115 course deliverables, we assess LLM performance using ChatGPT under a “minimal effort” protocol that simulates realistic student usage patterns. The investigation employs a rigorous testing methodology across multiple assessment formats, from auto-graded multiple choice questions to complex Python programming tasks and long-form analytical writing. Our analysis provides quantitative insights into AI’s strengths and limitations in handling mathematical formulations, coding challenges, and theoretical concepts in control systems engineering. The LLM achieved a B-grade performance (82.24%), approaching but not exceeding the class average (84.99%), with strongest results in structured assignments and greatest limitations in open-ended projects. The findings inform discussions about course design adaptation in response to AI advancement, moving beyond simple prohibition towards thoughtful integration of these tools in engineering education. Additional materials including syllabus, examination papers, design projects, and example responses can be found at the project website: https://gradegpt.github.io .

  PublisherDOI

• KnotDLO: Toward Interpretable Knot Tying

  ArXiv.org · 2025-06-27

  preprintOpen accessSenior author

  This work presents KnotDLO, a method for one-handed Deformable Linear Object (DLO) knot tying that is robust to occlusion, repeatable for varying rope initial configurations, interpretable for generating motion policies, and requires no human demonstrations or training. Grasp and target waypoints for future DLO states are planned from the current DLO shape. Grasp poses are computed from indexing the tracked piecewise linear curve representing the DLO state based on the current curve shape and are piecewise continuous. KnotDLO computes intermediate waypoints from the geometry of the current DLO state and the desired next state. The system decouples visual reasoning from control. In 16 trials of knot tying, KnotDLO achieves a 50% success rate in tying an overhand knot from previously unseen configurations.

  PublisherOA PDFDOI

• BOARD #127: WIP: Assessing aerospace students’ human-centered engineering design competency across multiple required courses

  2025-08-21

  article
  PublisherDOI

• Closed-loop Control of Polymerization Fronts during Frontal Polymerization of DCPD

  SSRN Electronic Journal · 2025-01-01

  preprintOpen accessSenior author
  PublisherDOI

• Efficient Extrinsic Self-Calibration of Multiple IMUs using Measurement Subset Selection

  arXiv (Cornell University) · 2024-07-02

  preprintOpen accessSenior author

  This paper addresses the problem of choosing a sparse subset of measurements for quick calibration parameter estimation. A standard solution to this is selecting a measurement only if its utility -- the difference between posterior (with the measurement) and prior information (without the measurement) -- exceeds some threshold. Theoretically, utility, a function of the parameter estimate, should be evaluated at the estimate obtained with all measurements selected so far, hence necessitating a recalibration with each new measurement. However, we hypothesize that utility is insensitive to changes in the parameter estimate for many systems of interest, suggesting that evaluating utility at some initial parameter guess would yield equivalent results in practice. We provide evidence supporting this hypothesis for extrinsic calibration of multiple inertial measurement units (IMUs), showing the reduction in calibration time by two orders of magnitude by forgoing recalibration for each measurement.

  PublisherOA PDFDOI

Recent grants

• RI: Small: Mechanics, Manipulation, and Perception of Deformable Objects for Robotic Manufacturing

  NSF · $450k · 2013–2017

• CAREER: Mechanics and Control of Brain-Machine Interface Systems

  NSF · $412k · 2010–2015

• CPS: Small: Mathematical, Computational, and Perceptual Foundations for Interactive Cyber-Physical Systems

  NSF · $540k · 2009–2013

Frequent coauthors

• Pagh Nissler

  University of Naples Federico II

  16 shared

• Drs Nancy

  Asia University

  16 shared

• Fausto Bordignon

  University of Bergamo

  16 shared

• Davide Brugali

  University of Bergamo

  16 shared

• Sébastien Chaves Gattaz

  University of Bergamo

  16 shared

• Cristiane Rúbia Ferreira

  National Institutes of Health

  16 shared

• Veronica Santos

  The Ohio State University

  16 shared

• Atsushi Kitazawa

  16 shared

Education

• Ph.D., Computer Science

  University of Illinois at Urbana-Champaign

  1994

• M.S., Computer Science

  University of Illinois at Urbana-Champaign

  1991

• B.S., Computer Science

  University of Illinois at Urbana-Champaign

  1989

Awards & honors

• Severns Faculty Scholar, University of Illinois at Urbana-Ch…