Determination of Operational Efficiency in Urban Public Transport Lines

Atalay, Ahmet · 2021 · Crossref

DOI: 10.36937/cebel.2021.001.004

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses the challenge of improving urban public transport efficiency amidst rising vehicle numbers, infrastructure deficits, and environmental concerns. The authors aim to classify bus lines based on operational performance to assist decision-makers in strategic planning. By identifying lines with high and low efficiency, the research seeks to optimize resource allocation, reduce energy consumption, and enhance service quality. The motivation stems from the need for data-driven tools to evaluate public transport systems, which are critical for sustainable urban mobility but often suffer from poor performance metrics that lead to economic subsidies and user dissatisfaction. The methodology employs clustering analysis to categorize 33 bus lines operated by the Erzurum Metropolitan Municipality and private companies in Erzurum, Turkey. The study utilizes functional data including capacity utilization rate (CUR), average passenger numbers (daily, peak hour, and weekend), vehicle counts, and distances traveled. Data was standardized using z-scores to handle varying magnitudes. The line G9 was excluded as an outlier due to significantly different operational values. Two clustering techniques were applied: the traditional k-means algorithm and the artificial intelligence-based Self-Organizing Mapping (SOM) neural network. The number of clusters was determined experimentally to be three, representing minimum, medium, and maximum performance levels. The results indicate that both k-means and SOM methods produced consistent classifications, dividing the remaining 32 lines into three distinct groups. Cluster 1 (Minimum) contained 8 lines, Cluster 2 (Medium) contained 14 lines, and Cluster 3 (Maximum) contained 10 lines. The analysis revealed that lines in the "big" or maximum cluster operate at or above full capacity, while those in the "small" or minimum cluster operate significantly under capacity. The medium cluster serves at levels near capacity. Additionally, the study found that lines connected to the Istasyon Referral and Administration Center exhibit density above capacity. The distance metrics between cluster centers confirmed clear separations between the groups, validating the robustness of the classification. The significance of this research lies in its demonstration that clustering analysis is an effective tool for evaluating public transport operational efficiency. The consistency between traditional statistical methods and neural network approaches suggests that SOM is a viable alternative for complex data classification in transportation planning. The findings provide actionable insights for urban planners to restructure routes, adjust fleet sizes, and improve service quality. By classifying lines according to functional efficiency, authorities can make informed decisions that yield economic savings and environmental benefits, ultimately enhancing the attractiveness of public transport over private vehicle use.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• regulatory evaluation