PhD researcher in robot control at the Automation and Control Institute (ACIN), TU Wien.
I work on model-based control of robots: multibody modelling, optimal control and their real-time implementation, with learned components where a model alone is not enough. My background runs from mechanical engineering through mobile robotics and perception to control of industrial and parallel robots.
PhD at ACIN, TU Wien (2026–), on skill-based control of industrial robots: a robot skill is specified as an optimal-control problem and executed by a model predictive controller together with a constraint layer at drive rate. Alongside the thesis I am building:
- a multibody dynamics library (C++ and Python) for serial, parallel and closed-chain mechanisms, validated against Pinocchio, MuJoCo and Drake;
- a real-time controller framework in C++ that runs the same controller in simulation and on Beckhoff TwinCAT;
- a MuJoCo digital twin of a linear-delta parallel robot, with MPC (acados, CasADi) and deployment through MATLAB/Simulink to TwinCAT.
This code is not public yet.
| project | |
|---|---|
| CA-MCW | Context-adaptive MPPI navigation: a small network re-weights the interpretable Nav2 critics each control cycle; feasibility checking stays classical. MSc thesis. |
| path-ilc | Path-parameter iterative learning control on a flexible-joint KUKA iiwa14 model in MuJoCo, with a learned layer for unseen paths and thermal drift. |
| pomona-nexus | ROS 2 workspace for an agricultural field robot: mapping, row navigation and UV-C treatment. |
| vehicle-manipulator-task-priority | Mobile base and arm controlled as one kinematic system with task-priority redundancy resolution. |
| pekf-slam-icp | Pose-based EKF SLAM with ICP scan matching. |
| lidar-exploration | Autonomous exploration with a 2D LiDAR, information-gain goals and Dubins-RRT*. |
robotics-learning: course-style modules (theory, C++, notebooks) on motion planning, state estimation, manipulator control, computer vision and sensor fusion.
- Context-Adaptive Multi-Critic Controller using Deep Learning-Based Cost Weighting in Autonomous Navigation. MSc thesis, 2026. PDF
- Path-Parameter ILC with a Learned Correction Layer for High-Accuracy Industrial Robots, 2026. PDF
- Kinematic Control of a Vehicle–Manipulator System Using Task-Priority Redundancy Resolution, 2025. PDF
- Pose-Based EKF SLAM Using ICP Laser Scan Matching, 2025. code
- Autonomous Exploration with a 2D LiDAR and Dubins-RRT* Planning, 2025. code
| area | |
|---|---|
| Modelling | rigid multibody dynamics (recursive algorithms, closed-chain constraints, reduced models), serial and parallel kinematics, flexible joints · Pinocchio, MuJoCo, Drake, SymPy, CasADi, Simscape |
| Control and optimisation | MPC and optimal control (acados, CasADi, IPOPT, qpOASES), MPPI, iterative learning control, computed-torque and impedance control, task-priority control, trajectory generation |
| Planning | graph search, sampling-based planners, Dubins paths, DWA, behaviour trees · MoveIt 2, Nav2 |
| Estimation and perception | Kalman and particle filters, EKF-SLAM, ICP, GNSS, point clouds, multi-view geometry · OpenCV |
| Learning | PyTorch, ONNX export to C++, PPO |
| Real-time and deployment | C++17, C99, real-time coding rules · Beckhoff TwinCAT 3, MATLAB/Simulink code generation · ROS 2, Docker |
| Engineering practice | CMake, GoogleTest, CI, sanitizers, Python, MATLAB, LaTeX, Git |
- PhD, Automation and Control Institute (ACIN), TU Wien, Austria. 2026–
- MSc Intelligent Field Robotic Systems, Erasmus Mundus Joint Master: Eötvös Loránd University, Hungary, and University of Girona, Spain. 2024–2026. Probabilistic robotics, manipulation, autonomous systems, multi-view geometry, 3D sensing and sensor fusion, deep learning. Thesis in industrial collaboration with European Knowledge Centre, Budapest.
- BEng Mechanical Engineering, Tribhuvan University, Nepal. 2017–2022. Robotics competitions: 1st place, TechFest IIT Bombay 2020.
- Research assistant at Kathmandu University and Tribhuvan University (2022–2023); lecturer at CTEVT, Nepal (2023–2024).
Languages: English (C1), Nepali (native), Hindi/Urdu, Spanish (A2), German (A1, learning).