Robotics and Autonomous Systems Lab

Robots that understand, adapt, and work alongside people.

We study intelligent robots for the real world — from companion robots that follow and assist humans, to agricultural robots in the field and legged robots that learn to walk.

Publications People
0Peer-reviewed papers
0Journal articles
0Conference papers
TSMCCOMPAGTHMST-ROIJRRRA-LTIITIETMECHICRAIROSAIMHumanoidsTSMCCOMPAGTHMST-ROIJRRRA-LTIITIETMECHICRAIROSAIMHumanoidsTSMCCOMPAGTHMST-ROIJRRRA-LTIITIETMECHICRAIROSAIMHumanoidsTSMCCOMPAGTHMST-ROIJRRRA-LTIITIETMECHICRAIROSAIMHumanoids
Research

What we work on

We combine learning, perception and control so that robots can act safely and usefully in dynamic, human-centered and unstructured environments.

01

Human–Robot Interaction

Companion robots with adaptive interaction strategies and proactive assistance, including multimodal language-model–driven care for the elderly.

  • Companion robots
  • Proactive assistance
  • Elderly care
02

Human-Aware Navigation

Socially compliant navigation, human-following and group-following robots that adapt to changing formations and crowded scenes.

  • Human/group following
  • Social navigation
  • Crowd navigation
03

Agricultural Robots

Autonomous navigation in orchards, traversability on uneven terrain, and context-aware pesticide spraying under changing weather.

  • Orchard navigation
  • Adaptive spraying
  • Uneven terrain
04

Legged Locomotion

Passive dynamic walking, kneed-walker guided policies, and humanoid robots that learn diverse whole-body skills.

  • Passive walking
  • Humanoids
  • Whole-body skills
05

Robot Learning

Deep reinforcement learning for navigation and multi-robot planning, and continual learning for robust human tracking.

  • Deep RL
  • Multi-robot planning
  • Continual learning
06

Vision–Language–Action

Vision–language models for scene understanding and decision making, and policy learning from synthetic video without real-world data.

  • Vision–language models
  • Scene reasoning
  • Synthetic video
News

Latest updates

View all
Oct2026

Excellent Doctoral Student Award

Cong-Thanh Vu received the Excellent Doctoral Student Award from National Cheng Kung University (NCKU), Taiwan.
Aug2026

Paper accepted to IEEE TSMC: Systems

Adaptive interaction strategies for human companion robots via deep reinforcement learning.
Jul2026

Paper published in COMPAG

Context-aware adaptive pesticide spraying for agricultural robots using vision–language models, Vol. 252.
Jun2026

Paper accepted to IEEE/RSJ IROS 2026

Adaptive companionship for group-following robots — see you in Pittsburgh, USA.
May2026

Three papers accepted to IEEE/ASME AIM 2026

Gimbal-based human tracking, VLM-driven elderly care, and humanoid learning from synthetic video — Genova, Italy.
Publications

Recent papers

View all