AI for Intelligent Transportation Systems: A Systematic Review of Applications in Demand-Responsive Transport

Area of Science:

• Transportation Science
• Artificial Intelligence
• Public Transit Operations

Background:

• Demand-Responsive Transport (DRT) offers flexible public transit solutions where fixed routes are inadequate.
• Machine Learning (ML) is increasingly explored for predictive analytics in DRT planning and operations.

Purpose of the Study:

• To systematically review ML-based predictive analytics in public transport DRT services.
• To assess the methodological quality, risk of bias, and applicability of existing ML research in DRT.

Main Methods:

• Systematic literature review of ML applications in DRT.
• Analysis of studies based on predictive task, data, models, validation, and metrics.
• Quality and bias assessment using a PROBAST+AI framework.

Main Results:

• Research on ML in DRT has grown rapidly, shifting towards integrated prediction and optimization frameworks.
• Significant heterogeneity exists in data, scales, models, and evaluation methods.
• High analysis-related risk of bias was noted due to insufficient independent validation and real-world testing.

Conclusions:

• Current ML research in DRT is promising but fragmented, requiring improved reproducibility and validation.
• Future research should focus on prospective, real-world assessments, robust comparisons, and interpretability for enhanced DRT.
• Addressing methodological gaps is crucial for advancing ML applications in public transport.