Date of Graduation

Fall 2025

Degree

Master of Science in Computer Science

Department

Computer Science

Committee Chair

Mukulika Ghosh

Abstract

Robotic navigation in dynamic environments presents significant challenges, particularly in managing interactions with moving obstacles while ensuring efficient path planning. Incorporating social navigation principles is crucial as robots share the same workspaces with humans frequently in daily life. This becomes essential for safe, efficient, and socially acceptable movements. In this thesis, I introduce a novel integration of social navigation strategies with topological path planning that leverages Discrete Morse Theory, a homotopical framework to enhance adaptability. My method dynamically assesses path feasibility and optimizes trajectory selection through three key strategies: waiting, deflection, and diverse path selection. Based on how close the robot is towards the goal, adaptive weighting is adopted to further balance between the strategies based on resource constraints and urgency, mimicking human-like decision-making. Experimental evaluations demonstrate an efficient solution with less than 100s in pre-processing computation overhead and improved path adaptability, resulting in a 30-60% increase in traversal time in environments containing 3-9 degrees of freedom robots and 15-60 dynamic obstacles. Further experimental evaluation shows that prioritizing waiting near the goal outperforms most of the environments in terms of traversal time and computation cost as the number of dynamic obstacles increases. This work is a step towards human-robot interactions with future applications in multi-robot planning and scheduling.

Keywords

path planning, motion planning, Vietoris-Rips complex, social navigation, discrete Morse theory, homotopy, dynamic obstacle interaction

Subject Categories

Artificial Intelligence and Robotics | Computer Sciences

Copyright

© S.M. Faiaz Mursalin

Download

Open Access

Share

COinS