Robotic Learning from Human
Reusable robot skills from human instructions, demonstrations and interaction history.
ROBOT LEARNING · MULTIMODAL INTERACTION · MANIPULATION
Exploring Intelligence.
Building with Robots.
I'm TAN RUNJIA, also known as Armine.
A PhD researcher at NTU, connecting
robot learning with real-world engineering.
Third-year PhD · Nanyang Technological University · Singapore
EXPLORE MY WORK ⟶
谭润家 TAN RUNJIA
RESEARCH · ENGINEERING
RESEARCH DIRECTIONS
Multimodal LLMs
Aligning language, gestures and visual perception to understand human intent.
Robotic Manipulation
Connecting perception, planning and control for precise, practical robot systems.
SELECTED RESEARCH & PROJECTS
01Research
DR-TSLM
Grounding “this” and “here” in speech and gestures, so robots can identify what to move and where it belongs.
02Official image · IntrinsicIndustry collaboration
Cable Assembly & Contact-Guided Insertion
Robotic cable insertion through visual registration and force feedback, in collaboration with Desay SV Singapore.
03Research & engineering
Vision-Servoed Microdevice Injection
A hardware-agnostic vision-servoed system for automated, precise injection of implantable microdevices.
04
VISUAL FEEDBACK · MANIPULATIONPerceive. Act. Refine.
Engineering
Desk Organization with Visual Feedback
Connecting visual-language perception and arm control to find, grasp and organize objects on a cluttered desk.
05Engineering
Mobile Manipulation & Language Interfaces
Integrating a robot arm, mobile base and vision sensors into a platform for natural-language-guided manipulation.
06Research
Long-Horizon Skill Learning
Turning instructions and interaction history into reusable skills for longer, compositional robot tasks.
INDUSTRY COLLABORATION
From the lab to assembly.
Working with Desay SV Singapore on robotic cable assembly, combining visual perception with contact-guided insertion.
SELECTED PUBLICATIONS
CIS-RAM 2026
DR-TSLM: Deictic Reference Resolution Via Temporal-Spatial Large Language Model for Human-Robot Interaction
ICRA 2026
An Autonomous and Hardware-Agnostic Vision-Servoed System for Microdevice Injection
IEEE Transactions on Industrial Informatics · 20(9), 11372–11383 · 2024
Personalized Disassembly Sequence Planning for a Human–Robot Hybrid Disassembly Cell
CIS-RAM 2024 · 14–19
Best PaperMulti-modal LLM-enabled Long-horizon Skill Learning for Robotic Manipulation
IEEE Robotics and Automation Letters · 7(4), 12451–12458 · 2022
Template-Based Category-Agnostic Instance Detection for Robotic Manipulation
† Equal contribution.
VIEW PUBLICATIONS ⟶THE JOURNEY
2020
B.Eng. in Automation
Hefei University of Technology2022
M.Sc. & visual perception research
Nanyang Technological University · RA-L2024
Skills & system integration
CIS-RAM Best Paper · IEEE TIINow · 2026
Third-year PhD researcher
ICRA · CIS-RAM · Industry collaboration✦ RECENT MILESTONES
AUG 2026 · CIS-RAMDR-TSLM: grounding words and gestures
⟶JUN 2026 · ICRAVision-servoed microdevice injection
⟶2026 · INDUSTRY COLLABORATIONMatrixX passes the Qualification Phase
⟶FROM THE BLOG ARCHIVE ⟶✦ LET'S CONNECT
I welcome conversations about robot learning, multimodal interaction, and bringing research into practical robotic systems.
✉ runjia.tan@ntu.edu.sg↗ GitHub / ar-mineSchool of Mechanical & Aerospace Engineering
Nanyang Technological University, Singapore