Manipulation
- Learning policies from demonstrations and datasets
- Generalization across objects, scenes, and tasks
- Benchmarking, sim-to-real transfer, and real-robot testing
George Washington University · Mechanical & Aerospace Engineering
Teaching
Courses connect physical behavior to mathematical models, then use simulation and experiments to test what those models predict.
Current courseFall 2026
MAE 4182
Electromechanical Control System Design
Modeling and feedback design for real electromechanical systems, emphasizing stability, performance, and robustness in time and frequency domains.
Visible derivations and interactive browser experiments help students connect equations to system response—and defend what each model predicts.
- Modeling
- Stability
- Robustness
- Controller design
MAE 6245Spring 2026 · Graduate course
Robotics Systems
A systems-level introduction to how robots sense, reason, and act under uncertainty.
Dynamics and control, sensing and vision, estimation, planning, and learning are studied as one connected pipeline.
View the Spring 2026 courseInside MAE 6245
Follow how errors propagate through an integrated robot.
Small modeling, sensing, and decision errors can compound through the pipeline. Students make assumptions explicit, reason about uncertainty, and evaluate system behavior rather than treating each module in isolation.
Navigation
- Offline and interactive navigation learning
- Robustness in new environments
- Comparison with classical planning baselines
The class is research-oriented, with critical reading, student-led paper discussions, and a semester-long open-ended project. Related project topics are welcome with instructor approval.
Selected course tools
From simulation to real-system evaluation.
Reference platforms used to explore robotics pipelines, learning, and localization.
01
Simulate
02