Objectives:
The goal of this research project is the elaboration of new, intelligent and flexible approaches to solving facilities layout problems in production management and logistics. It is planned to include in these algorithms, experts’ knowledge expressed similarly to a natural language. This can be achieved by employing linguistic patterns, which are based on fuzzy logic and linguistic expressions. They can be applied to defining relationships between objects and specifying criteria of the solutions’ quality. The project is focused on creating intelligent algorithms in three areas of facilities layout problems:
- Hierarchical problems. Layout optimization on multiple levels, e.g., on plant’s departments and department’s components.
- Multiple-criteria problems. Layout evaluation according to many measures.
- Flexible problems. Objects locations are not bound to predefined and constrained shapes or places.
The created approaches will undergo detailed series of analyses based on designed and conducted simulation experiments. They will aim at examining the usefulness of the proposed procedures in the domain of production management and logistics. The influence of algorithms’ parameters on the efficiency and effectiveness of finding solutions will also be investigated, along with performing sensitivity and robustness analyses. It is intended to compare the new proposals with other, intelligent methods.

Methodology:
The new algorithms will be developed in such programming environments as Delphi, Matlab, Python, C++.
It is planned to perform qualitative and quantitative analyses of the simulation studies. Qualitative assessment will involve comparisons of proposed approaches’ solutions with practical examples and experts’ recommendations. The quantitative assessments will be based on the experimental simulations outcomes. Verification of these quantitative results will be conducted by multi-factorial analyses of variance and generalized linear and nonlinear models. If appropriate assumptions for parametric tests are not met, their non-parametric equivalents may be employed. Post-hoc analyses will be used when applicable. Analyses of qualitative variables will be performed using contingency tables, loglinear analyses, and tested by the Chi square statistics.
Data manipulations and calculations will be carried out in Matlab, Statistica, IBM SPSS, or similar packages for advanced statistical and mathematical analyses.

Pioneering nature:
Application of multimodal logic and linguistic variables will make the proposed algorithms smarter in the sense of including experts’ knowledge in formal procedures. The increased flexibility will be achieved by developing new algorithms for hierarchical facilities layout problems where the optimisation would take place simultaneously on multiple levels of production or logistics management. The flexibility is related with utilizing scatter plots for facilities layout problems. They provide solutions for situations where target locations are not bound to predefined places. There are only a few methods offered in this respect, so new, intelligent approaches based on linguistic patterns and fuzzy logic are justifiable. As the proposals involve artificial intelligence aspects and take into account flexibility issues, the solutions obtained by these new procedures would probably be more adjusted to real, facilities layout problems.

Expected impact:
The successful accomplishment of this research project’s goal will provide new, intelligent tools for solving facilities layout problems occurring in production management and logistics. Proposed approaches will help to cope with data uncertainty by formal inclusion of the expert’s way of thinking, represented by linguistic patterns and fuzzy logic.
Simulations’ results will increase knowledge about the properties and capabilities of these algorithms which will facilitate their appropriate applications. This may contribute to the widespread use of various ways of solving problems associated with objects’ arrangements in practice.
The developed new approaches could be an inspiration for others in devising methods involving linguistic patterns and fuzzy logic. This research trend could be extended to versions of facilities layout problems with other than considered here assumptions and constraints. Furthermore, the new algorithms can be taken advantage of by other investigators as a possible components of prospect hybrid approaches