Case Study: ASTRI accelerates robotic manipulation development and cuts integration time 40% with MathWorks (MATLAB & Simulink)

ASTRI Accelerates Development of Robotic Manipulation System Using MBSE Digital Twin

ASTRI (Hong Kong Applied Science and Technology Research Institute) needed to accelerate development of robotic manipulation systems, cut rising development and production costs, and break down siloed engineering workflows that only revealed integration issues late in the process. To do this, ASTRI adopted a model-based systems engineering (MBSE) digital twin approach using MathWorks products including MATLAB, Simulink, Computer Vision Toolbox, Deep Learning Toolbox, and ROS Toolbox.

Using MathWorks tools, ASTRI built a digital twin for a robotic welding system that combined synthetic RGBD data, a trained deep neural network, ICP refinement, motion planning via ROS, and software-in-the-loop validation before deploying to hardware with code generation from MATLAB Coder and GPU Coder. The MathWorks-based solution reduced integration time by 40% and development time by 30%, enabled earlier issue resolution during design, and let distributed teams work in parallel to validate and deploy systems more quickly.

ASTRI

T. John Koo

ASTRI