Optimal Stacking (Phat $tax)

How would you pack your shopping bag? With the Phat $tax algorithm of course! We algorithmically generate valid packing arrangements and evaluate them using a value function.

Comment

Inspiration

Crappy packing by Coles

Automation is always the solution!

What it does

Algorithmically determine the optimal packing arrangement.

How we built it

We use a parallelised depth-first-search to generate all valid arrangements of items in a bag. We then use a multi-heuristic value function to determine the most optimal arrangement.

Challenges we ran into

Conceptualising a suitable idea. One of our earlier ideas was to improve recorded audio quality using machine learning was infeasible due to insufficient time to secure a suitable dataset.

Accomplishments that we're proud of

We bounced between a number ideas in the early stages and decided on algorithmic packing at noon on Saturday. With minimal time, we were able to learn some new things and implement a working prototype.

What we learned

  • Practical application of test-driven development process.
  • Learnt about parallel computing.

What's next for Optimal Stacking (Phat $tax)

  • Integrating value function within the generation process, allowing for alpha-beta pruning.
  • Improving the value function with additional/better heuristics.
  • Generalise algorithm to 3D.

Built With

Share this project:

Updates