Research team
Expertise
Macro-scale endeavors, primarily focused on AI-assisted monitoring and maintenance, providing solutions that enhance the reliability, efficiency, and sustainability of systems and infrastructure. Vision-Based Object Detection to develop algorithms for detecting and recognizing objects within visual data, often employing deep learning models for improved accuracy. Deep Learning Models to analyze complex datasets, extract meaningful patterns, and make predictions, contributing to enhanced decision-making processes. Time Series Analysis investigating temporal patterns within data to understand trends, fluctuations, and dependencies over time, particularly relevant for predictive maintenance and forecasting. Predictive Modeling that can forecast future outcomes or conditions based on historical data, aiding in proactive decision-making and resource optimization. Image Processing Techniques to enhance, analyze, and extract valuable information from visual data. This includes filtering, segmentation, and feature extraction for improved understanding. AI-Assisted Monitoring to continuously monitor and analyze real-time data streams, identifying anomalies or patterns that may require attention. Automation and Efficiency by Integrating AI and automation to improve efficiency in monitoring and maintenance processes, reducing manual intervention and enhancing overall system performance.
Exploring AI-guided Transportation Infrastructure Asset Management (AI-TIM).
Abstract
This project explores innovative approaches for AI-guided transportation infrastructure asset management through two complementary research paths: (i) health monitoring of road structural condition, and (ii) detection of erosion on cut-and-fill slopes. Both areas address pressing challenges in maintaining resilient transport systems and provide the foundation for future large-scale studies. For the road monitoring component, a field test section equipped with embedded sensors will be used to collect structural response data under real traffic and environmental conditions. The project will support the acquisition of additional hardware required for reliable and continuous data acquisition and remote monitoring. In the first phase, data collected for 3-4 months will be analyzed to refine AI-based processing techniques and assess the robustness of the data management process. In the second phase, the project will investigate alternative global sensing strategies that move beyond traditional point-based measurements. The goal is to conceptualize sensing approaches that can provide broader insights into structural performance and early signs of deterioration over the entire length of the road. The erosion component of the project focuses on the use of AI-enabled image analysis to identify and characterize erosion features on soil and rock slopes. Building on previous work on erosion processes, the study will explore the development of shape and texture metrics from image data. A central question is the suitability of different image sources for this task: whether satellite imagery provides sufficient resolution for erosion detection at scale, or whether higher-resolution UAV-based field imagery is required. The project will include the acquisition of relevant imagery and three months of focused methodological development. The outcomes of the project will include: • A functioning data acquisition and monitoring setup for a sensor-instrumented road section. • Preliminary analysis of local sensing data and conceptual exploration of global sensing approaches. • A set of candidate image-based metrics for erosion detection, tested on satellite and/or UAV imagery.Researcher(s)
- Promoter: Hernando David
- Co-promoter: Moins Ben
- Co-promoter: Ranyal Eshta
- Co-promoter: Van den bergh Wim
Research team(s)
Project type(s)
- Research Project
Enhancing Pavement Quality Through AI-Guided Compaction (EPAIC).
Abstract
Early damage on asphalt pavements is a high and avoidable social and economic cost. Research has shown that only 20% of early pavement failures are due to material defects, , while the remaining 80% can be attributed to the poor construction process itself. Currently these defects are only detected afterwards through core drilling and analysis. Too late, because the road surface has been realized. The innovative EPAIC project aims to the road construction industry in delivering more reliable and sustainable execution of the asphalt compaction process, which better meet established quality standards through the use of spatial and elevation monitoring technologies during road construction. EPAIC is introducing a digitised, AI-driven compaction system capable of continuously monitoring critical construction parameters, along with climatic conditions, during asphalt road construction. By analysing this data in real time using advanced AI algorithms, the system activates a smart alert mechanism that helps operators adjust key process variables during the process, such as the number, speed and frequency of rolling passes. As a result, optimal compaction quality is pursued. This proactive, real-time guidance helps prevent inferior quality on site, which currently can only be determined after the fact. EPAIC thus ensures an extended service life and no accelerated and unforeseen maintenance. EPAIC technology aims to become a market leader and addresses the persistent problem of an asphalt road that does not meet pre established quality standards, contributes to reduced infrastructure maintenance costs, a sustainable environment, reduced emissions and improved public health; factors that today's road construction industry is looking for. With a strong foundation of broad, technical expertise within the SuPAR Group, EPAIC is well positioned to move from concept to commercialisation and implement this technology in modern road construction.Researcher(s)
- Promoter: Omranian Seyed Reza
- Co-promoter: Ranyal Eshta
- Co-promoter: Van den bergh Wim
Research team(s)
Project type(s)
- Research Project