Congestion in multi-function parallel network DEA

Area of Science:

• Operations Research
• Industrial Engineering
• Economic Analysis

Background:

• Congestion, characterized by excessive inputs reducing outputs, increases costs and decreases efficiency.
• Existing Data Envelopment Analysis (DEA) methods primarily address decision-making units (DMUs) without network structures.
• Real-world applications often involve complex units with independent production subunits, necessitating specialized congestion detection.

Purpose of the Study:

• To propose a novel DEA model for identifying and evaluating congestion in multi-function parallel systems.
• To address the limitations of existing DEA models in handling network-structured DMUs.
• To provide an efficient computational approach for congestion analysis in complex production systems.

Main Methods:

• Development of a new DEA model tailored for multi-function parallel systems.
• Consideration of the operational aspects of individual subunits within each DMU.
• Comparative analysis of computational economy against existing 'black-box' methods.

Main Results:

• The proposed DEA model effectively identifies and evaluates congestion in multi-function parallel systems.
• The new method demonstrates significant computational advantages over traditional approaches.
• Validation through a numerical example and a real-world case study.

Conclusions:

• The developed DEA model offers a more effective and computationally economical solution for congestion analysis in complex production systems.
• This research extends DEA applications to network-structured DMUs, specifically multi-function parallel systems.
• The findings provide valuable insights for decision-makers aiming to optimize efficiency and reduce costs in intricate operational environments.