A summer project implementing A*, LPA*, and a homemade lazy variant of LPA* in Julia. Demo videos included in the 'videos' directory.
-
Make sure you have Julia installed
-
Create a 'frames' directory in the main directory. Static images will be saved here, and then combined into a video using ffmpeg.
-
To run A star: julia a_star.jl
-
To run lazy LPA star: julia lpa_star_lazy.jl
-
Videos are saved as a_star.mp4 and lpa_star.mp4 depending on which algorithm you run.
-
The simulation settings (start, goal, obstacles positions/shapes/speeds, etc) are located at the top of a_star.jl and lpa_star_lazy.jl.
- Implement a dynamic lookahead distance that changes based on how chaotic or unpredictable the environment is
- Incorporate side by side planning/execution/world dynamics using parallel computing in Julia
- Add more obstacle types and behaviors!
- Collision with an object currently just throws an error and stops planning.