Post by Nouran Mohamed
Mechatronics & Robotics Engineering Student @ Ain Shams University | President of Aviation Club | Specializing in Control Systems, IoT & Autonomous Technologies
Bringing Industrial Robotics to Life: A Comprehensive Digital Twin Project ! As part of our Mechatronics and Robotics Engineering coursework at Ain Shams University, my team and I recently modeled, simulated, and controlled a 6-DOF ABB IRB 120 industrial manipulator to execute a precise material handling pick-and-place task. To ensure absolute mathematical and physical accuracy, we didn't just rely on one approach. We evaluated the system through three distinct, parallel methodologies: 1️⃣ Analytical Modeling: Developing custom scripts from scratch for Forward/Inverse Kinematics and Euler-Lagrange dynamic formulations. 2️⃣ Numerical Simulation: Leveraging the MATLAB Robotics System Toolbox. 3️⃣ Physical Simulation: Deploying Computed Torque Control within Simulink and Simscape Multibody. By cross-validating the kinematic and dynamic outputs across all three methods, we successfully proved the structural integrity of our models to near-machine precision! My Specific Role: Bridging Theory and Simulation (Method 2) My main responsibility was driving the numerical models and optimizations using the MATLAB Robotics System Toolbox to connect our pure mathematical theory with dynamic simulation. My core technical contributions included: ▪️ Custom Architecture Construction: Programmatically building a custom rigidBodyTree object directly from verified Denavit-Hartenberg (DH) parameters to guarantee a strict 1:1 match with our custom mathematical models. ▪️ Trajectory Planning & IK: Generating Cartesian spatial paths for task waypoints and utilizing an optimization-based Inverse Kinematics solver to compute exact joint configurations. ▪️ Inverse Dynamics: Processing numerical derivatives through a recursive Newton-Euler formulation to compute feed-forward joint torques. ▪️ System Validation: Engineering an automated comparison loop to validate our numerical solver against the analytical models, confirming a maximum absolute difference of just 3.88e-16. Check out the attached video to see the full pick-and-place project simulation in action! I’d also like to thank my teammates—Ibrahim El-Dawy, Mostafa Abdelhamid, and sara shaker—for their solid work on the analytical modeling, and Simscape simulations. It was a great collaborative effort. #Robotics #Mechatronics #MATLAB #IndustrialAutomation #DigitalTwin #ABB #Kinematics #AinShamsUniversity #Engineering
Video Content