Cycling represents a significant share of urban transportation, especially in terms of last-mile delivery. It has clear benefits for delivery times, as well as for environmental issues related to freight distribution. Furthermore, the increasing accessibility of low-cost environmental sensors (LCSs) provides an opportunity for urban monitoring in any situation. Moving in this direction, this research aims to investigate the use of LCSs to monitor particulate PM2.5 and PM10 levels and map them over delivery ride paths. The calibration process took 49 days of measurements into account, spanning different seasonal conditions (from May 2024 to November 2024). The employment of multiple linear regression and robust regression revealed a strong correlation between pollutant levels from two sources and other factors such as temperature and humidity. Subsequently, a one-month trial was carried out in the city of Faenza (Italy). In this study, a commercially available LCS was mounted on a cargo bike for measurement during delivery processes. This approach was adopted to reveal biker exposure to air pollutants. In this way, the operator’s route would be designed to select the best route in terms of delivery timing and polluting factors in the future. Furthermore, the integration of environmental monitoring to map urban environments has the potential to enhance the accuracy of local pollution mapping, thereby supporting municipal efforts to inform citizens and develop targeted air quality strategies.

This work aims to investigate the effectiveness of road maintenance interventions by analyzing changes in the International Roughness Index (IRI) by means of crowdsourced connected vehicle data. For this purpose, 136 pavement maintenance interventions on a single lane were considered over a period between 2021 and 2024. A multiple linear regression model (R² = 0.780) has been employed as statistical tool to assess the relationship between pre/post-intervention IRI scores and various factors. Despite the fact that results showed a general improvement in pavement condition, the effectiveness of the interventions was found to be influenced by several factors. In particular, intervention on the middle lane appears to be the most effective for enhancing section characteristics, and the effectiveness of maintenance on the overall condition of the section tends to be reduced as the number of lanes increases. Additionally, maintenance appears to be less effective in improving post-maintenance performance as the initial IRI value increases; this suggests that severely deteriorated sections may require more extensive rehabilitation strategies. The ultimate aim of study is to provide evidence to support the inclusion of crowdsource vehicle data in Pavement Management Systems (PMSs) to optimize maintenance planning and resource allocation.

The forecasting of road surface conditions is a pivotal component for intelligent transportation systems, in terms of supporting maintenance planning, safety and mobility management. The increasing availability of large-scale monitoring data, collected from passenger vehicle fleets, enables the development of data-driven forecasting approaches. However, systematic comparisons between classical time-series models and machine-learning methods in this context remain limited. The proposed benchmarking framework evaluates direct road surface roughness forecasts at 1-, 7-, 14-, 30-, and 90-day horizons using multi-year vehicle-derived data collected across heterogeneous road segments. Daily roughness indicators are derived from raw measurements and modeled following a consistent, segment-wise experimental protocol. The proposed analysis involves the evaluation of multiple machine-learning regressors including Ridge, Random Forest and Gradient Boosting which are trained on lagged observations and rolling statistics. Performance of the models is assessed using two error metrics: unweighted and uncertainty-aware weighted. Findings indicate significant variations in predictive accuracy and robustness across models and segments, emphasizing the influence of feature-based learning strategies and data-quality weighting. The research provides a scalable and transparent methodology for evaluating forecasting models on vehicle-based road monitoring data, contributing practical guidance for the deployment of artificial intelligence in Intelligent Transport Systems (ITSs).