Dynamic Infrastructure and University of Bath Test AI Approaches for Infrastructure

NEW YORK and BATH, England, Sept. 30, 2026 /CourierPR/ -- Dynamic Infrastructure, the leading Engineering AI platform for critical civil assets, has partnered with the University of Bath to explore the performance of various AI approaches in real-world infrastructure tasks. This collaboration aims to address a critical gap in the industry by testing how well different AI models can generalize to infrastructure they have never seen before.

The study, led by Arik Voronov, AI R&D Lead at Dynamic Infrastructure, and Professor Thomas Kjeldsen from the University of Bath's Department of Architecture and Civil Engineering, focuses on using computer vision to identify infrastructure conditions from inspection images, particularly blockage and obstruction in drainage assets. The research is part of MSc research by Arline Osorio Moreno and seeks to understand how well these models perform across different sites, which can vary significantly in construction, camera equipment, and inspection practices.

Dynamic Infrastructure manages thousands of structures across 15 states and countries, providing autonomous risk intelligence to predict deterioration and assess structural health. The company's platform replaces subjective opinions with data-driven insights, helping prevent catastrophic failures. By collaborating with the University of Bath, the company aims to develop more robust and reliable AI tools for infrastructure management.

The research compares several AI approaches, including vision-language models, vision foundation models, supervised deep learning, and classical machine-learning methods. Each approach has its own strengths and weaknesses, such as accuracy, robustness, and the amount of labeled data required. Early results indicate that while broad models like vision-language models have significant capabilities, they do not yet provide the robustness needed for reliable real-world implementation. More specialized approaches, on the other hand, can perform substantially better.

The study's findings will contribute to the development of more reliable and consistent AI tools for infrastructure management. The partners intend to continue their collaboration and aim to develop the work into a peer-reviewed publication.

The University of Bath, one of the UK's leading universities, is recognized for its high-impact research and excellence in education. It is ranked among the top 10% of universities globally, placing 125th in the QS World University Rankings 2027. The university has achieved a triple Gold award in the last Teaching Excellence Framework 2023, the highest awards possible, for both the overall assessment and for student outcomes and experience.

This project highlights the ongoing efforts to bridge the gap between academic research and practical applications in the field of engineering, ensuring that new AI tools are both scientifically rigorous and capable of delivering tangible benefits for infrastructure operators and wider society.