Infrastructures

Gravel roads are essential for rural accessibility but are increasingly vulnerable to climate-induced flooding, erosion, sedimentation, and slope instability. Conventional engineering approaches often rely on costly hard infrastructure with limited environmental benefits. Nature-Based Solutions (NbSs) and sustainable natural materials offer complementary approaches for enhancing road resilience through improved hydrological performance, soil stabilisation, and ecosystem restoration. This systematic review synthesised evidence on the application of NbSs and sustainable natural materials for climate-resilient gravel roads in flood-prone regions. Following PRISMA 2020 guidelines, the literature published between 2000 and 2026 was systematically screened, resulting in 18 high-quality studies for critical synthesis. Four principal intervention categories emerged: vegetation and soil bioengineering, biodegradable erosion-control materials, road-water harvesting, and catchment-scale drainage and land-cover management. Findings demonstrate that these interventions can reduce runoff, erosion, sediment yield, and slope instability while improving water availability, ecosystem restoration, and rural accessibility. Greater resilience benefits were associated with integrating NbSs and conventional engineering through hybrid green–grey infrastructure and adaptive maintenance. Based on the synthesis, an Integrated Nature-Based Climate-Resilient Gravel-Road Framework is proposed, linking climate-risk assessment, ecosystem-based interventions, sustainable materials, road-water management, and performance monitoring to support sustainable and climate-resilient rural transport infrastructure.

Although mechanistic–empirical approaches have gained increasing prominence in pavement design, the evaluation and selection of soil–aggregate mixtures still frequently rely on gradation criteria and the California Bearing Ratio (CBR), which may not fully capture the behavior of these materials under repeated loading, particularly in the case of tropical lateritic soils. In this context, this study aimed to evaluate the use of permanent deformation obtained from multistage repeated-load triaxial testing as a performance-oriented tool for comparing and ranking soil–aggregate compositions. Three mixtures composed of lateritic soil and basalt aggregate (SA50-50, SA40-60, and SA30-70) were proportioned to comply with the same reference gradation envelope and evaluated through CBR, resilient modulus, and multistage permanent deformation tests. These tests were complemented by a single-stage permanent deformation test conducted under the same stress state for all three mixtures. Increasing the aggregate content increased the CBR from 61% to 144%, whereas the average resilient modulus ranged from 301 to 356 MPa, without showing a consistent increasing trend. SA40-60 exhibited the lowest permanent deformation levels along all three multistage loading paths, with final permanent strains of approximately 0.85%, 0.81%, and 2.12% for TC50, TC70, and TC100, respectively. The single-stage test showed the same overall trend as the multistage tests, with a final permanent deformation of 1.09 mm for SA40-60, compared with 2.17 mm and 3.52 mm for SA50-50 and SA30-70, respectively. The results indicate that conventional parameters and resilient response, when considered individually, do not fully describe the behavior of these mixtures with respect to the accumulation of irreversible deformation. Multistage permanent deformation testing therefore proved to be a useful tool for comparing different soil–aggregate mixtures and supporting the selection of the best-performing composition among those evaluated.

This study examined infrastructure-related casualty involvement in police-reported crashes and assessed how record quality and crash context affect interpretation. The cohort comprised 6190 Eslamshahr crashes recorded from March 2022 to June 2024, including 1805 injury or fatal crashes. Models developed on 5299 records were assessed on 891 spring-1403 records. The reference CatBoost achieved ROC-AUC 0.857, average precision 0.725 and Brier score 0.120; development-only sigmoid recalibration reduced Brier to 0.112. A 39-configuration comparison did not improve CatBoost and yielded stronger average-precision and Brier results for random forest on the same existing test. Adjusted associations for inadequate nighttime lighting (OR 2.78, 95% CI 1.66–4.64) and visibility obstacles (OR 1.89, 1.05–3.38) remained elevated across width and influence checks. Exactly 12 m accounted for 64.3% of widths and materially affected geometric estimates. Removing crash type and at-fault vehicle information reduced CatBoost ROC-AUC to 0.532. The findings support retrospective examination of recorded lighting and visibility conditions and show why documentation, predictor availability and calibration matter when interpreting severity models.