Journal of Transportation Systems

Comparative Analysis of Machine Learning Algorithms for Road Traffic Forecasting in Developing Countries

Maloani Saidi Georges — 2026-08-13

Accurate short-term traffic forecasting is essential for effective urban mobility management, yet developing countries face unique challenges due to sparse, fragmented, and frequently missing data. This systematic review compares four major machine learning families: classical statistical models (ARIMA), shallow machine learning (Support Vector Regression and Random Forest), deep recurrent networks (LSTM and GRU), and spatio-temporal graph models (Graph WaveNet and STGCN) for road traffic prediction. A total of 42 peer-reviewed studies published between 2018 and 2025 were synthesized. Unlike traditional benchmarks that prioritize accuracy alone, this review evaluates models across three operational dimensions: predictive accuracy, robustness to data missingness (up to 40%), and computational efficiency (training time and memory footprint).

Results reveal a fundamental trilemma: graph models achieve the highest accuracy with a mean MAPE of approximately 6.3%, but degrade severely under missing data and require over 45 minutes per training epoch. In contrast, ARIMA is highly robust and fast, yet inaccurate with a MAPE of 11.4%. LSTM and GRU offer the most balanced compromise, with GRU achieving a MAPE of 7.9% in under 19 minutes.

This study provides a context-aware decision framework, recommending GRU with light imputation as the pragmatic choice for low-resource settings, and calls for standardized benchmarks tailored to data-sparse environments.

Waste-Plastic Strips for Reinforcing Pavement Subgrade Soil: Mechanisms, Performance, Sustainability and Research Needs

Nikhil Kushwaha — 2026-08-11

Randomly distributed strips cut from post-consumer plastic offer a potential method to improve weak pavement subgrades while creating an engineered reuse route for difficult-to-recycle films, pouches, and bottles. This review critically synthesizes the mechanisms, material variables, laboratory performance, pavement implications, and environmental safeguards associated with strip-reinforced soil. The evidence shows that improvement arises from interface friction, edge bearing, pull-out resistance, tensile bridging, lateral confinement, and crack interception. These mechanisms are not activated uniformly; their contribution depends on polymer stiffness and surface condition, strip length, width, thickness and aspect ratio, dosage by dry soil mass, soil gradation and plasticity, moisture state, compaction energy, and mixing quality. Across published studies, moderate plastic contents commonly increase California bearing ratio (CBR), unconfined compressive strength, shear strength, stiffness, and post-peak ductility. However, the response is usually non-monotonic because excessive strip content displaces mineral particles, lowers dry density, encourages folding and clustering, and creates locally weak or poorly compacted zones. In the benchmark dataset examined in this review, cleaned low-density polyethene strips measuring 5 mm × 30 mm (aspect ratio 6) at 3% of dry soil mass produced the best common laboratory condition for three soils. Soaked CBR increased from 4.25%, 5.82% and 3.40% to 8.46%, 12.74% and 6.28%, respectively, while maximum dry density remained largely retained. Nevertheless, this optimum is material- and process-specific and should not be generalized without replicate statistics and complete geometry–dosage matrices. A practical design framework is therefore proposed that treats strength, compactability, constructability, durability, and environmental containment as simultaneous acceptance criteria. Before routine field adoption, the technology requires repeated-load and rutting tests, wet–dry and temperature cycling, abrasion and leaching assessment, microplastic-release controls, quality-controlled mixing procedures, and monitored trial pavement sections.

Techno-Economic Feasibility Study of the Proposed Pune Outer Ring Road

Satyam M. Thakur — 2026-05-12

The Pune Metropolitan Region is experiencing acute urban mobility constraints attributable to the convergence of three National Highways — NH-48, NH-65, and NH-60 — within its core, resulting in average arterial speeds below 15 km/h during peak hours and escalating Vehicle Operating Costs (VOC). This paper presents a comprehensive techno-economic feasibility study of the proposed 174 km Pune Outer Ring Road (PORR), a greenfield expressway designed to divert inter-city commercial freight from the urban core. The study addresses three principal dimensions: structural and geotechnical design, risk classification, and financial viability. Geotechnical challenges posed by the Deccan Traps formation are mitigated through the adoption of the New Austrian Tunneling Method (NATM) across six tunnel segments totaling 3.655 km, complemented by deep-pile foundations for river crossings over the Mula, Mutha, and Indrayani rivers. A structured risk assessment matrix identifies and addresses geotechnical variability, ecological sensitivity, infrastructure conflicts, and land acquisition complexities across 67 peri-urban villages. The proposed ₹42,000 crore capital expenditure is structured under a Design-Build-Finance-Operate-Transfer (DBFOT) procurement model, with post-construction monetization through Toll-Operate-Transfer (TOT) leases and Infrastructure Investment Trusts (InvITs), supplemented by Land Value Capture (LVC) mechanisms along the reserved 110 m right-of-way corridor. Traffic modelling projections indicate a diversion of over 25,000 PCUs on key segments, with total motorized vehicle volumes forecast to reach 270,065 PCU by 2040. The project yields an Economic Internal Rate of Return (EIRR) exceeding 18%, affirming its long-term economic viability and its potential as a replicable model for mega-infrastructure financing in rapidly urbanizing Indian metropolitan regions.

Project Governance and Performance in Federal Road Infrastructure: A Study of Nepal's Strategic Road Network

Vivek Kumar Jha — 2026-08-12

Federal road projects in Nepal consistently experience delays, cost overruns, and implementation failures despite established governance frameworks. This study examines the relationship between project governance mechanisms and performance outcomes in Nepal’s federal road construction sector. Using a mixed-methods approach combining Relative Importance Index (RII) analysis of survey data from 30 construction professionals, key informant interviews with six project stakeholders, and document review of three federal road projects, the research identifies critical governance gaps affecting project performance. Findings reveal that poor inter-agency coordination (RII = 0.853), inadequate grievance redress mechanisms (RII = 0.787), and political interference (RII = 0.653) are the most significant barriers to effective project governance. The study contributes to the literature on infrastructure governance in developing countries by demonstrating how institutional fragmentation — exacerbated by Nepal’s federal transition — undermines project performance. Policy recommendations include establishing dedicated coordination mechanisms, strengthening grievance redress systems, and enhancing institutional capacity within the Department of Roads.

A Hybrid RII-AHP Framework for Prioritizing Stakeholder Engagement in Transportation Infrastructure Projects: A Case Study of Nepal's Federal Road Sector

Vivek Kumar Jha — 2026-08-06

Effective stakeholder engagement is critical for the successful delivery of transportation infrastructure projects, yet many government agencies lack structured frameworks for prioritizing engagement activities. This study develops a practical Stakeholder Engagement and Environmental and Social (E&S) Disclosure Framework for federal road projects in Nepal. Using a mixed-methods approach, the research combines the Relative Importance Index (RII) and the Analytic Hierarchy Process (AHP) to prioritize framework components based on input from 30 construction professionals and 16 experts. RII analysis reveals that Regular Public Consultation and Feedback Collection (RII = 0.773) and Timely Information Disclosure (RII = 0.733) are the highest-priority framework components. AHP analysis demonstrates that Timely Information Disclosure vs. Grievance Redress carries the highest decision weight (37.7%), indicating that stakeholders view information transparency and dispute resolution as the most critical governance functions. The proposed framework provides a replicable tool for transportation agencies in developing countries seeking to institutionalize stakeholder engagement practices.