Research notes · 2026
Planning paths through changing worlds.
Notes on a research direction developing around robotics and path-planning algorithms.
Taha Burak Özdemir · İnönü University · Department of Computer Engineering
A direction in formation
My current research direction is taking shape around robotics and path planning: how an autonomous system can represent its environment, search for a viable route, and revise that route when the world changes.
I am currently reviewing established methods, implementing small comparisons, and refining the doctoral question through those experiments.
Questions taking shape
Three connected questions currently organize the work.
Representation and search.
Which environmental representation and search strategy gives the right balance of efficiency, clarity, and route quality?
Changing constraints.
How should a planner respond when obstacles, costs, or task priorities change after planning has begun?
Evaluation.
Which benchmarks and reporting practices make improvements reproducible and meaningful beyond a single simulation?
The foundation
My master's-era research used machine learning and image analysis for ship detection in Synthetic Aperture Radar imagery. That work produced the publications listed on this site and built practical experience in noisy data, perception pipelines, and experimental evaluation.
SAR now sits in the publication record rather than at the center of my current work. It remains useful experience in data preparation, algorithm comparison, and experimental evaluation.
Adjacent systems questions
Autonomous and multi-agent systems remain a natural extension of planning. I am particularly interested in how route planning, task allocation, and coordination interact when agents have different capabilities or incomplete information.
Current working approach
Start with a baseline.
Reimplement established methods before comparing changes or extensions.
Record the assumptions.
Keep the environment, constraints, and evaluation measures explicit.
Use small experiments.
Keep implementations inspectable while the research question is still becoming more specific.
Open to
Research conversations around robotics, path planning, autonomous systems, and experimental evaluation.
get in touch