Shaoshuai Mou

About

Shaoshuai Mou is a robotics and control researcher whose work sits at the intersection of optimal control theory, machine learning, and human-robot interaction. He is best known for pioneering advances in **inverse optimal control (IOC)**, developing rigorous methodologies that allow systems to infer an agent's underlying objective function from observed trajectory data — even when those observations are incomplete, fragmentary, or sparse. A central thread of Mou's research is the **Pontryagin Differentiable Programming (PDP)** framework, an end-to-end learning and control paradigm that differentiates through Pontryagin's Maximum Principle to unify a broad class of learning and control problems. Building on this foundation, his Continuous PDP method enables robots to learn from just a handful of sparse keyframe demonstrations, dramatically reducing the data burden in robot learning. His work on learning from human directional corrections further advances intuitive human-robot collaboration by removing the need for precisely calibrated correction magnitudes. Collectively, Mou's papers have accumulated over 200 citations, with his foundational IOC and PDP works each exceeding 50 citations. His contributions span multi-robot mission planning and iterative IOC methods, reflecting a broad commitment to making autonomous systems more adaptive, data-efficient, and responsive to human intent.

Research Focus

Key Achievements

Top Papers

1. 1
   Inverse optimal control from incomplete trajectory observations

   52 citations · 2021
2. 2
   Inverse Optimal Control for Multiphase Cost Functions

   51 citations · 2019
3. 3
   Learning From Sparse Demonstrations

   32 citations · 2022
4. 4
   Pontryagin Differentiable Programming: An End-to-End Learning and\n Control Framework

   29 citations · 2019
5. 5
   Learning From Human Directional Corrections

   18 citations · 2022
6. 6
   Inverse Optimal Control with Incomplete Observations.

   14 citations · 2018
7. 7
   Real-Time Multi-Robot Mission Planning in Cluttered Environment

   6 citations · 2024
8. 8
   An Iterative Method for Inverse Optimal Control

   5 citations · 2022
9. 9
   Learning from Sparse Demonstrations

   5 citations · 2020
10. 10
    Inverse Optimal Control from Demonstration Segments.

    3 citations · 2020

Key Collaborators

Contact & Links