Fully Web and Virtual Reality Based Collaborative System for Controlling Multi Robots Industrial Environment

Abstract

Industry 4.0 is accelerating the integration of robotic systems into industrial environments, yet direct interaction between novice workers and robots raises significant safety concerns, making robotic simulation an essential training tool prior to real deployment. Existing platforms, however, face a persistent trade-off : browser-based solutions offer accessibility but sacrifice physical realism, while high-fidelity simulators require complex local installations and substantial computing resources. This thesis presents the design, implementation, and validation of a fully browser-based collaborative robotic simulation platform that bridges this gap, built on a five-layer modular architecture integrating the Cannon.js physics engine and the Three.js rendering library to deliver 60 Hz rigid-body simulation — including gravity, collision detection, and joint constraint solving — directly in a standard web browser with no installation required, complemented by an integrated dual-camera perception system that provides real-time object tracking and proximity-based collision avoidance via Frustum culling. Experimental validation using a three-tier strategy confirms a 90% success rate for autonomous Pick & Place cycles, collision resolution rates between 95% and 100%, multi-user WebSocket synchronisation at 28 ms average latency for four concurrent participants, and a rendering performance of 54 FPS for four simultaneous robots ; these results were obtained under a limited experimental scope, restricted to box-shaped objects within a single warehouse environment and evaluated on CPU-only hardware, which future work should extend to irregular objects, diverse industrial settings, and GPU-accelerated configurations. Overall, this work establishes the feasibility of combining web accessibility, real-time physics, intelligent grasping, and collaborative multi-robot interaction within a single browser-based platform, offering a validated and reproducible foundation for the development of next-generation Industry 5.0 collaborative robotic systems