Published in IEEE International Conference on Unmanned Aircraft Systems (ICUAS), 2018
Force-control formulation for cooperative aerial manipulation, regulating interaction forces between multiple UAVs and a shared payload.
Recommended citation: Thapa S., Bai H., Acosta J.A. "Force Control in Cooperative Aerial Manipulation." IEEE International Conference on Unmanned Aircraft Systems (ICUAS), 2018.
Published in IFAC Workshop on Networked & Autonomous Air & Space Systems (NAASS), 2018
Force-controlled cooperative aerial load transport, presented at the IFAC NAASS workshop.
Recommended citation: Thapa S., Bai H., Acosta J.A. "Cooperative Aerial Load Transport with Force Control." IFAC Workshop on Networked & Autonomous Air & Space Systems (NAASS), 2018.
Published in American Control Conference (ACC), 2018
Cooperative control for multi-UAV aerial load transport with simultaneous payload attitude stabilization.
Recommended citation: Thapa S., Bai H., Acosta J.A. "Cooperative Aerial Load Transport with Attitude Stabilization." American Control Conference (ACC), 2018.
Published in IEEE Control Systems Letters (L-CSS), with CDC presentation option, 2019
Concurrent-learning adaptive control for a multi-robot team transporting an unknown payload, enabling real-time estimation of payload mass and drag with guaranteed parameter convergence under relaxed persistent-excitation requirements.
Recommended citation: Thapa S., Self R., Bai H., Kamalapurkar R. "Cooperative Manipulation of an Unknown Payload with Concurrent Mass and Drag Force Estimation." IEEE Control Systems Letters (L-CSS), 2019. https://ieeexplore.ieee.org/abstract/document/8725526
Published in Journal of Intelligent & Robotic Systems (JINT), 2020
Decentralized adaptive force-consensus controllers enabling a multi-UAV team to jointly transport a payload without direct inter-robot communication under unknown payload mass and external disturbances, with Lyapunov-based stability analysis.
Recommended citation: Thapa S., Bai H., Acosta J.A. "Cooperative Aerial Manipulation with Decentralized Adaptive Force-Consensus Control." Journal of Intelligent & Robotic Systems (JINT), 2020. https://link.springer.com/article/10.1007/s10846-019-01048-4
Published in In preparation for arXiv, 2026
A comparative tutorial benchmarking PID, state-dependent LQR, geometric SE(3), and nonlinear adaptive controllers for quadrotor trajectory tracking, validated in Flightmare, Gazebo, and ROS.
Recommended citation: Thapa S. "A Comparative Tutorial on Autonomous Quadrotor Trajectory Tracking Control: PID, State-Dependent LQR, Geometric SE(3), and Nonlinear Adaptive Control." In preparation for arXiv.
Published in In preparation for the American Control Conference (ACC) 2027, 2026
An event-triggered PID in which the hold is a state of a finite state machine rather than a compensation term, so the integral term is held whenever the command is held and no maximum sampling period is required. The hold band is calibrated from the actuator deadband and the sensor resolution measured in a single constant-command field test, requiring no plant model. On the published benchmarks of Durand and Marchand the controller reproduces the reference update counts exactly and matches the hybrid event-based PID at identical gains, with a further reduction in updates under measurement noise. Includes a field deployment of the code-generated controller on production hardware.
Recommended citation: Thapa S. "A Practical State-Machine Based Event PID Controller." In preparation for the American Control Conference (ACC) 2027.
Published in In preparation for the American Control Conference (ACC) 2027, 2026
A closed-form feasibility certificate for the single-trial reverse parking maneuver, reducing the existence of a curvature-continuous solution from a given entry pose to the sign of a scalar quadratic discriminant — an O(1) test with no iteration or root-finding, and equivalently a line–circle intersection. Evaluated backward from the parked pose, the certificate defines the set of viable entry poses, and that set becomes the goal region of a forward hybrid-A* search, so the global-to-local handoff is feasible by construction rather than by chance. The local planner re-solves from the pose actually reached, absorbing grid-snap error. Validated in a Monte Carlo study against tracked, collision-checked outcomes, with a nonlinear path-frame law executing the maneuver in reverse under steering saturation.
Recommended citation: Thapa S. "Certified Handoff Between Global Search and Local Planning for Automated Parallel Parking." In preparation for the American Control Conference (ACC) 2027.
Undergraduate course, School of Mechanical & Aerospace Engineering, Oklahoma State University, 2016
Teaching assistant for Elementary Dynamics (ENSC 2123): led discussion sessions, delivered partial lectures, and graded coursework.
Undergraduate course, School of Mechanical & Aerospace Engineering, Oklahoma State University, 2018
Teaching assistant for Mechatronics Design (MAE 4733): lab instruction and grading for a class of 28 students.