Game AI · Algorithms
Unity AI Navigation Algorithms
Shared Unity environment comparing Ad Hoc rules, A* pathfinding and Monte Carlo Tree Search through three autonomous agents and a live score display.
- Unity 2022.3
- C#
- A*
- Monte Carlo Tree Search
- Rule-Based AI
01
Comparison Context
The project places rule-based, heuristic-search and simulation-based agents in one low-poly Unity environment and exposes their score progression through a shared HUD.
- Algorithms
- 3
- Environment
- Unity
- Demo
- 00:40
02
Shared Test Environment
All three agents operate in the same low-poly scene with visible characters, targets and environmental obstacles. A shared HUD reports each score during execution.
- 01Ad Hoc navigation driven by explicit rules
- 02A* navigation driven by heuristic path search
- 03Monte Carlo Tree Search navigation driven by simulated decision branches
- 04Shared targets, obstacles, animated agents and live score display
03
Three Approaches
Ad Hoc
Rule-based decisions
A*
Heuristic path search
Monte Carlo Tree Search
Simulation-based decisions
04
Recorded Run
- A* score at 00:35
- 90
- Ad Hoc score at 00:35
- 70
- MCTS score at 00:35
- 10
These values are a snapshot from the supplied recording. They describe this run only and should not be interpreted as a general performance ranking.
05
Live Demonstration
The short recording is the strongest evidence for the comparison because it shows the agents moving while all three scores update in the same view.
06
Engineering Approach
- 01Unity 2022.3 scene and C# agent scripts
- 02The same environment and score interface make the three behaviors observable side by side
- 03The recording captures score evolution during one 40-second execution
07
Limitations
- 01The recorded scores represent one visible run, not a statistically controlled benchmark.
- 02Runtime, path length, explored nodes and success-rate measurements are not documented in the supplied media.
08