Projects

// PROJECT_DATABASE

23 projects, 6 domains, 2021 → now

Research, control, locomotion, industry deployments, drones, and embedded hardware. Hover any row to expand. Click to dive in.

23PROJECTS

6DOMAINS

5WITH PAPERS

12CORE TECH

// FILTER · DOMAIN click to filter

ALL 23 RESEARCH 6 LOCOMOTION 2 CONTROL 4 INDUSTRY 5 DRONES 3 HARDWARE 3

// ALL_PROJECTS · INDEX

1. 01 INDUSTRY

   Digital Twin · Motoman MH180 Handling Robot · ISAAC SIM

   2026 IN PROGRESS ↗

   Physics-accurate digital twin of a Yaskawa Motoman MH180 handling robot in NVIDIA Isaac Sim. Sub-micron Lula-IK path tracking, resolved-rate teleoperation, rigid pick-and-carry, and live WebRTC streaming so the virtual robot can be operated from anywhere.

   Isaac SimLula IKRMPflowTeleoperationPick & CarryWebRTC
2. 02 INDUSTRY

   Weld Automation using Yaskawa Robots · AR2010 · ARC WELDING

   2026 IN PROGRESS ↗

   Automatic arc-welding motion generation for car-body sub-assemblies. Two Yaskawa AR2010 robots weld seams on a jig-held part — torch trajectory, orientation, and approach generated from CAD geometry to ≤ 0.1 mm, validated in MuJoCo + Isaac Sim and deployed to a real YRC1000.

   Arc WeldingTrajectory GenerationTorch OrientationIKIsaac SimYaskawa
3. 03 INDUSTRY

   Tensiq · TT · TACTILE AI

   2026 SHIPPING ↗

   Universal tactile intelligence for robotics. A standardized tensor format and processing pipeline that turns any tactile sensor into model-ready, physics-grounded data — the “MP4 for touch”.

   Tactile AIMulti-SensorPipelineRobot LearningStandardization
4. 04 RESEARCH

   SUREFlow · IROS '26

   2026 PUBLISHED ↗

   State-space uncertainty-aware residual flow matching for robust robot manipulation. Mamba-backbone VLA — 92.6% LIBERO success rate with just 179M parameters, beats MaIL by +34.3%.

   Flow MatchingMambaVLAUncertaintyLIBERO
5. 05 CONTROL

   RL Prescribed Performance · FTC · IEEE TCYBE

   2026 UNDER REVIEW ↗

   Reinforcement-learning-based neuro-optimal control for robot manipulators under composite actuator faults. Predefined-time tracking via prescribed performance functions + actor–critic–identifier RL. Submitted to IEEE Transactions on Cybernetics.

   RLNeuro-Optimal ControlFault TolerancePPFManipulator
6. 06 RESEARCH

   OpenARM-VLA · OSS · REAZON

   2026 OSS CONTRIB ↗

   Open-source contribution to Reazon Research · OpenArm. A VLA framework on the OpenArm platform in Isaac Sim. Head-to-head benchmark of MambaVLA (state-space) vs MDT Transformer with an RL-teacher synthetic data pipeline.

   VLAMambaMDTIsaac SimOpenArmOSS
7. 07 RESEARCH

   MambaVLA

   2026 CCNC ’26 ↗

   Scalable, efficient Vision-Language-Action transformer with state-space architecture for robotic manipulation. Mamba SSM at the policy core.

   Mamba SSMVLAPyTorchManipulation
8. 08 LOCOMOTION

   QROOT

   2025 NeurIPS ’25 ↗

   Integrated diffusion transformer + reinforcement learning for quadrupedal locomotion. End-to-end gait synthesis from a latent prior.

   DiTRLQuadrupedIsaac Sim
9. 09 RESEARCH

   DiffDAIL

   2025 UNDER REVIEW ↗

   Diffusion-enhanced, vision-guided imitation learning with discrete latent representations. Robust policies from noisy demonstrations.

   DiffusionImitation LearningLatentPyTorch
10. 10 CONTROL

    DLDMP

    2025 NODYCON ’25 ↗

    Discrete latent diffusion motion planning for manipulators.

    DiffusionMotion PlanningManipulator
11. 11 INDUSTRY

    Tissue Processing

    2024 DEPLOYED ↗

    Vision-guided tissue manipulation. 85.7% autonomous success on ViperX arms at Dexweaver.

    ViperXACTDexweaver
12. 12 RESEARCH

    ACT on Franka

    2024 IEMEK ’24 ↗

    Action Chunking Transformer implementation for vision-guided pick-and-place on the Franka arm.

    ACTFrankaPick & Place
13. 13 INDUSTRY

    Autonomous Harvesting

    2024 DEPLOYED ↗

    Cucumber harvesting using YOLOv11 detection with ViperX arms on a mobile AGV.

    YOLOv11ViperXAGV
14. 14 RESEARCH

    Imitation × Diffusion · Franka

    2024 ↗

    Combining imitation learning with diffusion processes on the Franka manipulator.

    ImitationDiffusionFranka
15. 15 CONTROL

    Hybrid MPC + ILC

    2024 KNU-EERC ’24 ↗

    Hybrid model predictive and iterative learning control for leader-follower robotic tracking.

    MPCILCTracking
16. 16 LOCOMOTION

    Reinforcement Learning · Basics

    2023 ↗

    Fundamental RL algorithms — DQN through SAC — implemented end-to-end.

    PPOSACGymnasium
17. 17 CONTROL

    DNN-NMPC for CSTR

    2023 ↗

    Deep neural network–based nonlinear MPC for a continuous stirred-tank reactor.

    DNNNMPCCSTR
18. 18 DRONES

    Tello + Object Detection

    2023 ↗

    Tello drone teleop with keyboard and onboard object detection.

    TelloDetectionUAV
19. 19 DRONES

    Drone Control · Keyboard

    2023 ↗

    Keyboard-based teleop for UAV navigation and waypoint following.

    UAVTeleopPython
20. 20 DRONES

    Drone Trajectory · LPV-MPC

    2023 ↗

    UAV trajectory tracking with linear parameter-varying MPC.

    LPV-MPCUAVPython
21. 21 HARDWARE

    32-bit MIPS CPU

    2022 ↗

    32-bit pipelined MIPS CPU built from scratch in Verilog.

    VerilogMIPSPipeline
22. 22 HARDWARE

    Smart Home · IoT

    2021 ↗

    ESP32-based home automation with cloud control.

    ESP32IoTCloud
23. 23 HARDWARE

    Verilog Digital Designs

    2021 ↗

    Collection of digital design projects in Verilog HDL.

    VerilogHDLDigital