Evolutionary Soft Robot Design

Project Overview

Created a physics-based simulation environment to evolve soft robots that learn efficient locomotion patterns. The project focused on using genetic algorithms to optimize the movement parameters of a snake-like robot constructed from masses and springs.

Custom Physics Engine Simulation

Technical Implementation

• Built custom physics engine from scratch incorporating:
  • Gravity simulation
  • Damping effects
  • Friction modeling
• Designed snake-like robot using mass-spring system
• Implemented evolutionary algorithm to optimize:
  • Spring constants (K)
  • Oscillation frequency (omega)
  • Phase differences (phi)

Project Videos

📹 Evolution Process and Final Results

📹 Robot Movement Demonstration

📹 Final Demonstration

Project Images

Snake-like Robot Configuration with Mass-Spring System

Optimization Results Showing Movement Evolution