HUMAN AVOIDANCE BY MOBILE PLATFORM

Abstract

The increasing demand for autonomous mobile robots has led to their widespread use across various industries, including healthcare, logistics, and manufacturing. Effective human avoidance is crucial for safe navigation, necessitating sophisticated collision-avoidance techniques. This project focuses on the development and implementation of a Model Predictive Control (MPC) algorithm to enhance mobile robot navigation in dynamic environments, specifically addressing the challenge of human avoidance. Unlike traditional obstacle avoidance techniques, which often rely on reactive measures, MPC offers a predictive approach that allows robots to anticipate human movement and adjust their trajectory in real time. The project is implemented on jackal robot with LiDAR and RGB-D camera using Clearpath existing sources. The project involves integrating a customized MPC controller into the existing Nav2 framework, as well as incorporating human detection YOLOv8 package. By forecasting human movement and incorporating constraints related to human-robot interaction, this project aims to create safer, more efficient robots capable of operating in complex, human-populated environments across various industries.