Skip to content

astra-uu-se/COMPD

Repository files navigation

COMPD

License

Constraint Optimization of MicroPlate Designs

This is a constraint optimization model written in MiniZinc. It is used for effective microplate layout generation and improves on PLAID in several crucial ways:

  • improved performance on larger microplates and dramatically decreased resource consumption,
  • COMPD generates more effective layouts on larger microplates
  • COMPD supports more microplate templates:
    • plate lines with an odd number of rows and/or columns,
    • forcing corners to be empty
  • flexible approach to spreading the materials of the same type, i.e., the thresholds for minimum possible distances are not hard-coded and, instead, are found dynamically.

The full explanation can be found in the paper "Constraint Optimization of MicroPlate Designs" (TBD: link and DOI)

To test, you can use the *.dzn files from evaluation/regression-tests directory or a small example from https://github.com/pharmbio/plaid/blob/main/small-example.dzn

GUI

To make the workflow with COMPD more fluid and intuitive, you can use MPLACE - an external GUI utility developed in Python.

Evaluation

The directory evaluation contains instructions on how to execute the evaluation section of the paper. Note that we use version 1.0 in the evaluation (as it was updated since the article submission). Version 1.0 is available here.

Version history

See file VERSION_HISTORY.md for details

Credits

COMPD is developed by Ramiz Gindullin and Maria Andreina Francisco Rodriguez.

The following from AAAI'26 can be used to cite COMPD:

Gindullin, R., & Rodríguez, M. A. F. (2026, March). Constraint Optimization of MicroPlate Designs. In Proceedings of the AAAI Conference on Artificial Intelligence (Vol. 40, No. 17, pp. 14243-14250).

@inproceedings{gindullin2026constraint,
  title={Constraint Optimization of MicroPlate Designs},
  author={Gindullin, Ramiz and Rodr{\'\i}guez, Mar{\'\i}a Andre{\'\i}na Francisco},
  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
  volume={40},
  number={17},
  pages={14243--14250},
  year={2026}
}

License

COMPD has an Apache 2.0 LICENSE. The COMPD team accepts no responsibility or liability for the use of COMPD or any direct or indirect damages arising out of its use.

About

Constraint Optimization of MicroPlate Designs

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors