Skip to content

avocadoxD0/RnD-Project

Repository files navigation

Arm-to-Arm Bilateral Control of Robotic Manipulators

Overview

This R&D project presents the design, implementation, and validation of a joint-space bilateral control framework for robotic manipulators. The system enables real-time motion synchronization and force reflection between a master and a slave robotic arm, allowing one robot to perceive interaction forces experienced by the other.

The project uses the Interbotix Reactor X150 as the master manipulator and the Interbotix Viper X300 as the slave. The bilateral control loop was first validated in simulation using ROS2, RViz, and PyBullet, followed by partial deployment on physical hardware to study real-world limitations.


Objectives

  • Implement arm-to-arm bilateral control using joint-space mapping
  • Enable force reflection from the slave to the master during environmental interaction
  • Study transparency and stability characteristics of bilateral control
  • Compare simulation performance with real hardware behavior

System Architecture

The bilateral control system consists of two coupled information flows:

  • Forward Path (Master → Slave):
    Master joint angles are mapped to the slave using a predefined joint-space mapping matrix.

  • Feedback Path (Slave → Master):
    Reaction torques generated due to environment interaction are reflected back to the master.

A compliant admittance control model is implemented on the master side to convert reflected torques into smooth positional offsets when direct torque control is unavailable.


Mathematical Framework

The control formulation follows standard rigid-body joint-space dynamics:

  • Joint-space mapping between manipulators with mismatched DOF
  • Torque computation using Jacobian-transformed interaction forces
  • Gain-scaled force reflection to the master
  • Admittance-based compliance for stable force feedback

Stability is addressed using passivity-based considerations, gain tuning, torque filtering, and dead-zone handling to prevent oscillations.


Simulation & Validation

  • Initial master–slave mapping verified in RViz using Interbotix URDF models
  • Full bilateral control loop implemented and tested in PyBullet
  • Simulation demonstrated:
    • Stable joint synchronization
    • Correct force-feedback activation during contact
    • Robust response under obstacle interaction

Simulation results confirmed the effectiveness of joint-space bilateral control under idealized conditions.


Hardware Deployment

The bilateral control framework was deployed on the physical X150–X300 robotic arm pair. While partial bilateral motion was achieved, sustained synchronized operation was limited by:

  • Control-loop timing inconsistencies
  • Python execution and communication latency
  • Actuator thermal and hardware constraints
  • Noise and uncertainty in torque estimation

These observations highlight the gap between simulation and real-world implementation.


Key Observations

  • Joint-space bilateral control performs reliably in simulation
  • Force reflection behaves predictably during environment contact
  • Hardware deployment exposes challenges related to timing, actuation, and sensing
  • Real-world constraints significantly affect bilateral control stability

Tech Stack

  • Middleware: ROS2
  • Simulation: PyBullet
  • Visualization: RViz
  • Robotic Arms: Interbotix Reactor X150, Interbotix Viper X300
  • Actuators: Dynamixel servo motors
  • Programming Language: Python
  • Control Techniques: Bilateral control, admittance control, joint-space mapping

Future Work

  • Integration of full IK/FK for Cartesian-space bilateral control
  • Latency-aware and high-frequency control loops
  • Improved force sensing and torque estimation
  • Enhanced hardware reliability and thermal management
  • Extension to real-world teleoperation tasks

Author

Farhaan Nasir
R&D Project – Robotics & Control Systems


License

This project is intended for academic and research purposes.

About

Bilateral Control of Robotic Arms

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages