Joonho Lee, Ph.D.
Sr. Engineer, Guidance Navigation Control & Autonomy
Boeing Research and Technology
Everett, Washington 98204, USA
Research interests
Intent estimator design for non-cooperative objects using Machine Learning and Inverse Reinforcement Learning
Planning algorithm design based on Machine Learning, Inverse Reinforcement Learning, and Deep Neural Networks
2D and 3D tracker design for the vision detection system based on Machine Learning
Sensor fusion algorithm design for ADS-B signals and vision detects
Robust Deep Reinforcement Learning: Two-time Scale Actor-Critic based on Stochastic Approximation (projects: highway merge, multi-actuator control designs for a self-driving car, Landing an UAV on a moving vessel)
Deep Learning: Graph Neural Networks (projects: a vision based trailer sway estimator using deep learning and dynamic models)
Nonlinear Controls: Multi-Time-Scale Approaches, High-Gain Observer Design, On-line Optimizations (Dynamic Inversion), etc;
Machine Learning: Gaussian Process (estimations, probabilistic behavior planning (Factor Graph), etc)
Stability Analysis for Stochastic Systems Using Stochastic Approximations
Dynamics: Under-actuated Mechanical Systems - Vehicle Dynamics (Chassis, Wheels controls - Traction Control System), Helicopters, Quadrotors, Robotic Boats, and a Powered Prosthetic Leg, etc