Mohamed Amine Hajji / Projects

Fuzzy Logic Obstacle Avoidance for a Mobile Robot

A from-scratch Mamdani fuzzy inference engine — reproducing a published robotics paper — steering a simulated robot around obstacles in a live browser visualization.

Paper (DOI)

This project reproduces "Design and Implementation of Fuzzy Logic for Obstacle Avoidance in Differential Drive Mobile Robot" (Puriyanto & Mustofa, 2024): a robot that senses its surroundings with proximity sensors and uses a Mamdani fuzzy inference system — implemented entirely from scratch, no fuzzy-logic library — to decide how to steer. It ships as a real-time browser simulator instead of physical hardware: a Flask/Flask-SocketIO backend runs the fuzzy controller and a 2D ray-casting sensor simulation, streaming robot state to an HTML5 Canvas front end at ~30 updates/sec with live membership-function and rule-activation charts alongside the simulation.

Highlights

• Full Mamdani fuzzy inference engine from scratch — fuzzification, MIN-rule evaluation, MAX aggregation, centroid defuzzification — zero fuzzy-logic library dependencies.
• 19 heuristic rules across priority tiers (emergency avoidance, caution, wall-following, clear path) driving continuous steering-angle and speed outputs.
• Four selectable test environments (random obstacles, corridor, U-shaped trap, empty space) plus a collision counter for gauging navigation performance.

Stack: Python, Flask, Socket.IO, NumPy, JavaScript