Tsun-Hsuan (Johnson) Wang
Tsun-Hsuan (Johnson) Wang is a PhD candidate whose research integrates computer science with artificial intelligence and machine learning to create real-world autonomous systems, such as robot cars, and to explore foundational questions in robot control. With the support of a MathWorks Fellowship, Johnson will expand his innovative work to create high-fidelity environments and sim-to- real approaches for control learning that will safely expose agents to challenging and dangerous edge cases. The goal of this research is to develop robot systems that can learn from prior experiences yet are robust enough to guarantee safety under real-world conditions and untrained scenarios. Johnson recently developed and implemented VISTA, a system based on this line of research that is now freely available to the research community. MATLAB supports vital aspects of Johnson’s work, chiefly developing algorithms in simulated environments and generating data for policy learning and robot deployment. His work has already yielded valuable contributions to autonomous driving and has the potential to broadly advance machine learning for safety-critical applications in many industries.