Yao-Ting (Daniel) Huang

yah032@ucsd.edu

I’m Yao-Ting (Daniel) Huang, an ECE M.S. student at UC San Diego (ISRC) developing visuotactile robot policies for dexterous, contact-rich manipulation. Before joining UCSD, I completed my B.S. in Electrical Engineering at National Taiwan University, where I conducted research in the NTU Robot Learning Lab under Prof. Shau-Hua Sun and in the Advanced Control Lab under Prof. Li-Chen Fu. My undergraduate studies explored unsupervised reinforcement learning and vision-based UAV navigation, giving me early experience in algorithmic design, perception, and control for autonomous systems. These projects laid the foundation for my current work at UCSD, where I integrate tactile feedback, imitation learning, and sim-to-real transfer to advance robot dexterity.

Research focus

Visuotactile policy learning for dexterous manipulation under contact-rich dynamics
Multimodal fusion of vision, force and tactile feedback for robust slip and contact-state estimation, enabling data-efficient imitation learning and sim-to-real transfer for surgical and in-hand tasks
End-to-end teleoperation system development (firmware, communication, simulation setup, force-feedback computation, and kinematics modelling)

Recent work

During my first year at UCSD, I joined the ARCLab advised by Prof. Michael Yip and worked closely with Soofiyan Atar. As a Graduate Researcher, I developed the N2D haptic glove teleoperation system, building its complete hardware-software pipeline—including firmware, communication protocols, Isaac Sim integration, force-feedback computation, and kinematic modeling. This system enables precise bidirectional teleoperation between a human operator and a dexterous robotic hand for contact-rich surgical and manipulation tasks. Apart from that, I collect visuotactile demonstrations to train imitation policies that generalize across simulation and real-world settings. Two related works are currently under review for ICRA 2026.

news

Sep 15, 2025	Submitted 2 papers to ICRA’26.
Jun 16, 2025	Started SRIP(Summer Research Internship Program) in ARCLab
Sep 23, 2024	I started my journey of pursuing ECE master degree at UCSD!.
Aug 14, 2023	Started intern in Industrial Technology Research Institute self-driving automobile group.
Jun 11, 2023	Graduated from National Taiwan University, majoring in Electrical Engineering.

selected publications

In-Hand Manipulation of Articulated Tools with Dexterous Robot Hands with Sim-to-Real Transfer

Soofiyan Atar, Daniel Huang, Florian Richter, and Michael Yip

arXiv preprint arXiv:2509.23075, 2025

Abs HTML PDF

Reinforcement learning (RL) and sim-to-real transfer have advanced robotic manipulation of rigid objects. Yet, policies remain brittle when applied to articulated mechanisms due to contact-rich dynamics and under-modeled joint phenomena such as friction, stiction, backlash, and clearances. We address this challenge through dexterous in-hand manipulation of articulated tools using a robotic hand with reduced articulation and kinematic redundancy relative to the human hand. Our controller augments a simulation-trained base policy with a sensor-driven refinement learned from hardware demonstrations, conditioning on proprioception and target articulation states while fusing whole-hand tactile and force feedback with the policy’s internal action intent via cross-attention-based integration. This design enables online adaptation to instance-specific articulation properties, stabilizes contact interactions, regulates internal forces, and coordinates coupled-link motion under perturbations. We validate our approach across a diversity of real-world examples, including scissors, pliers, minimally invasive surgical tools, and staplers. We achieve robust transfer from simulation to hardware, improved disturbance resilience, and generalization to previously unseen articulated tools, thereby reducing reliance on precise physical modeling in contact-rich settings.
The N2D Haptic Glove: A Multi-Finger Glove for 2D Directional Force Feedback for Contact Rich Manipulation

Yao-Ting Huang, Kaitlin Calimbahin, Jake Honma, Logan Li, Omar Hernandezand, and Michael Yip

2025

Abs PDF

Humans rely on directional fingertip cues, particularly normal forces, to probe and regulate contact during manipulation, yet most wearable gloves render only vibration or single-axis force, leaving direction ambiguous. Without directional cues, users must infer contact force from vision alone, which leads to over-pressing, inconsistent control, and reduced precision in robotic teleoperation. We present the N2D Haptic Glove, a multi-finger wearable device that renders planar vector forces at each fingertip using a compact, transparency-focused mechanical design with on-hand actuation that preserves natural motion. Through benchtop validations and teleoperation user studies involving haptically teleoperating a humanoid robot, we demonstrate that multi-directional fingertip feedback significantly improves user ability to regulate contact, reduce overshoot, and improve consistency in interactions compared to visual-only or 1D haptic feedback. These findings establish the N2D Haptic Glove and directional finger-based haptics device as an important modality for contact-rich teleoperation, immersive simulation, and robot learning from demonstrations.