hero-image
HOME
hero-image

Bilateral Teleoperation of 1-DOF System with Kinesthetic Feedback

hero-image
Roy Suliman

Project Timeline

Jan 2024 - Sep-2024

OVERVIEW

I co-developed a bilateral teleoperation system that enables real-time kinesthetic feedback between two 1-DOF systems—one industry-grade (Quanser linear cart) and one custom-built robotic arm (“TeleBot”). The project aimed to allow an operator to control a remote device and feel the interaction forces from the remote side, enabling natural, responsive motion for remote manipulation tasks. The system was implemented in two configurations: 1. Quanser-to-Quanser Setup: Two linear carts controlled via Simulink using UDP over WiFi. 2. Custom Arm-to-Arm Setup: Two ESP32-based arms communicating via WiFi using embedded C control algorithms. This work required integrating mechanical, electrical, and control systems while achieving low latency (<200 ms) and stable force-position control.

HighlightS

  • Co-developed control algorithm for bilateral teleoperation and kinesthetic feedback
  • Implemented wireless connectivity for remote operation beyond lab settings
  • Designed, 3D printed, and assembled all mechanical components for the custom TeleBot arm
  • Developed complete CAD models for both the arm and demonstration setup
  • Managed all Quanser-related tasks: Simulink model development, testing, and integration
  • Achieved ±1° accuracy in displacement tracking and <200 ms latency in feedback loop
  • Presented working prototype at Capstone Tradeshow, receiving recognition for potential applications in hazardous environments, industrial automation, and remote operations.

SKILLS

SimulinkMATLABUDP/TCP NetworkingReal-Time SystemsESP32DC Motor ControlSignal FilteringSystem IntegrationMechanical Design (CAD)Prototyping3D PrintingControl TheoryC Programming (Embedded)

Additional Details

System Description
The bilateral teleoperation system consisted of a master and a slave device, each capable of commanding the other and providing real-time kinesthetic feedback.

Phase I – Quanser Implementation:

  • Two independent Simulink models ran on separate PCs connected to Q2-USB interfaces.

  • Systems communicated via UDP protocol over Ethernet/WiFi.

  • Communication was optimized to minimize jitter and ensure natural force feedback.

Phase II – Custom Arm Implementation (TeleBot):

  • Each ESP32 controlled a DC motor and read encoder/potentiometer data.

  • Force-position control implemented using interrupt-driven C code.

  • Arms were built from laser-cut acrylic, belt-driven for smooth motion.

  • Wireless control achieved with <200 ms delay.

  • System allowed either arm to serve as master or slave dynamically.

Key Challenges & Solutions:

  • Voltage noise: Mitigated via improved wiring and filtering.

  • Signal drift: Addressed through calibration routines.

  • Latency: Reduced through optimization of communication protocol and code execution.

Budget Summary (Table Reference):
Table 2 – Budget Breakdown: Total project cost: $483.97. Major expenses included motors, ESP32 boards, Wi-Fi adapters, and 3D printing materials. (Insert table screenshot here with caption: “Component-level budget summary for the TeleBot project.”)

Demonstration:
The final integrated system was demonstrated at the Capstone Tradeshow on July 26, 2024, where attendees could operate the master device and feel remote interaction forces in real time.

Figures to Include:

  1. Figure – Phase I Setup: Two Quanser carts connected via Ethernet during testing.

  2. Figure – Phase II Custom Arm: Custom TeleBot arm with ESP32, belt-drive, and encoder integration.

  3. Figure – System Overview Diagram: Block diagram showing communication, control loops, and master-slave relationship.

Future Improvements:

  • Integrate force sensors for tactile feedback

  • Increase degrees of freedom for more complex tasks

  • Explore alternative microcontrollers (STM32, Raspberry Pi)

  • Use brushless motors for smoother, quieter operation

  • Upgrade encoders and feedback systems for higher precision

  • Test alternative communication protocols (MQTT, CAN Bus)

 

 

The bilateral teleoperation system consisted of a "master" and a "slave" device that could each command the other and respond with kinesthetic feedback. Initially, we used two Quanser setups (each including a linear cart and encoder, motor driver, and Q2-USB interface). We developed independent Simulink models on two separate PCs to run in real-time, communicating via UDP protocol. Communication and synchronization were optimized to reduce latency and jitter, allowing natural interaction between systems.


In the second phase, we designed our own hardware system using ESP32 microcontrollers. Each ESP32 controlled a DC motor and read from a potentiometer or encoder, transmitting real-time data to its counterpart over WiFi. The system supported seamless switching of master/slave roles, allowing either arm to lead.


We faced and overcame challenges related to voltage noise, current limitations, and signal drift. Through multiple iterations, we improved control accuracy and response time. The mechanical arms were built from laser-cut acrylic parts and featured belt-driven mechanisms for smooth motion.


This project not only demonstrated our technical skills in mechatronics but also our ability to work in a multidisciplinary team, communicate effectively, and build robust systems from concept to demonstration.

lowinertia
Portfolio Builder for Engineers
Created by Aram Lee
© 2025 Low Inertia. All rights reserved.