A Visual Debugger for Pathfinding Search

This work aims to develop a visualisation tool to help practitioners better understand existing pathfinding algorithms and to help researchers more easily implement new pathfinding algorithms. The tool is intended to be used as both a pedagogical teaching aid and as a useful and online debugger for researchers. It can be motivated as follows:

• Practitioners interested in existing pathfinding algorithms (e.g., students) typically seek to understand such methods by studying diagrammatic examples, algorithmic pseudo-code or fragments of actual source code if an implementation is available. Each of these approaches is limited and, from a pedagogical perspective, poor substitutes for a hands-on interactive application.
• Researchers interested in pathfinding algorithms typically proceed by developing code from scratch, iterating a write-run-test programmatic loop. Errors in logic are often difficult to find because mistakes/bugs can arise in the middle of solving a complicated instance. The typical approach is to debug the algorithm by analysing many thousands lines of text output. This approach is time consuming and itself error prone.