Abstract Manager

In 2025, Brun-Way Highways Operations Inc (BHOI) marked its 20^th year of operating, maintaining, and rehabilitation 275 km of highway infrastructure in New Brunswick, extending from the Quebec border to Fredericton. Over these two decades, BHOI has built a comprehensive pavement performance database that includes IRI, rutting, SDI and periodic core and base-material sampling. This long-term dataset now provides a rare opportunity to evaluate rehabilitation strategies with confidence supported by empirical evidence.

When BHOI first assumed responsibility for the highway, several rehabilitation interventions were undertaken without the ability to reliably forecast performance outcomes. Today, the availability of a rich longitudinal dataset has transformed our decision-making process, enabling us to identify rehabilitation practices that consistently deliver durable results.

Using machine learning techniques, we analyzed a large dataset from the U.S. Federal Highway Administration’s Long-Term Pavement Performance Program alongside BHOI’s historical data. This analysis allowed us to identify the most influential factors affecting pavement deterioration, including climatic conditions, asphalt layer thickness, and average annual daily truck traffic. We then refined our approach by developing deterioration curves tailored specifically to our highway network, modeled at the lane level. A second machine learning analysis focused on the effectiveness of various rehabilitation treatments, allowing us to quantify how each intervention influences pavement performance over time. Combined, these two lines of analysis form the basis of our predictive model, which now guides long-term rehabilitation planning by producing reliable, performance-based forecasts.

This predictive capability enhances BHOI’s ability to meet contractual performance requirements, optimize infrastructure investments, and adapt strategies to New Brunswick’s specific environmental and traffic conditions. In 2026, prior to the conference, BHOI will compare newly collected pavement performance data to the model’s predictions to further validate and refine our approach.

Looking ahead, BHOI intends to develop similar predictive models for our other highway operations across Canada, supporting more efficient investment planning and the delivery of long-lasting pavement infrastructure for the benefit of road users nationwide.