Physics-Informed AI for Infrastructure Resilience

Introduction 

Civil infrastructure management is currently facing a significant transition. Highly detailed static models in BIM and GIS are now accompanied by a rapidly growing influx of raw, dynamic data from IoT systems and structural health monitoring sensors. However, enabling effective integration and communication between these two data environments remains a persistent challenge. A persistent limitation in current research is that many AI applications treat complex physical assets as generic data sets, often overlooking the underlying principles of structural mechanics, geotechnics, and material degradation.

This Special Issue explores what happens when core engineering logic informs and constrains the application of AI. Submissions are invited to investigate how established digital tools, such as BIM, GIS, and digital twins, can be used to impose physically meaningful constraints on machine learning models. The ultimate aim is to move past generalised "big data" enthusiasm towards robust, physics-informed AI methods that can genuinely support practical decision-making and lifecycle asset management. Unlike conventional studies that treat infrastructure systems as abstract data environments, this Special Issue encourages contributions in which engineering knowledge actively shapes AI model development, enabling more reliable, interpretable, and deployable solutions.

The originality of this Special Issue lies in its direct response to a persistent divide within current research. At present, the field is largely split into two domains: studies focused on developing complex infrastructure data structures and ontologies, and studies focused exclusively on applying deep learning algorithms to raw sensor data. There remains surprisingly limited integration between these structural frameworks and applied analytics. This Special Issue is uniquely positioned to bridge this gap by advancing a hybrid approach in which structured knowledge directly enhances machine learning models. By investigating how semantic representations of infrastructure systems can guide AI development for civil infrastructures, the issue promotes a novel direction. Ensuring that domain-specific civil engineering principles inform AI model developments is an underexplored, yet essential, step toward their safe deployment at scale.

The need for this Special Issue is further driven by an immediate industry imperative: the transition toward resilient, net-zero infrastructure. Asset managers face increasing pressure to reduce maintenance costs, track embodied carbon, predict long-term degradation, and consider climate resilience under extreme conditions. Conventional software systems struggle to handle this multidimensional complexity, accelerating the sector's adoption of artificial intelligence. However, without appropriate engineering constraints, many AI-based solutions risk limited reliability and trust. Consequently, the development of domain-constrained digital tools capable of large-scale lifecycle analysis has become a key research priority. This Special Issue captures this momentum, by focusing on physics-informed approaches that are robust, transparent and applicable in real-world infrastructure systems

List of topic areas

• Integration of GIS and BIM (infra-BIM) and digital twins for managing large-scale infrastructure systems
• Application of open standards for data interoperability and lifecycle asset management
• Development of digital frameworks for real-world lifecycle assessments (LCA) and embodied carbon tracking in civil infrastructure
• Physics-informed and engineering-constrained machine learning for structural health monitoring (SHM)
• Integration of IoT sensor networks with infrastructure digital twins for predictive maintenance and real-time monitoring
• Application of computer vision, LiDAR, and robotic systems for autonomous asset inspection and on-site defect detection
• Utilisation of natural language processing (NLP) and semantic web technologies for compliance checking and regulatory analysis
• AI-driven decision-support tools for balancing maintenance efforts, cost, structural safety, and carbon performance
• AI and digital twin applications for climate adaptation, vulnerability assessment, disaster response, and infrastructure resilience.

Submissions Information

Submissions are made using ScholarOne Manuscripts. Registration and access are available at:  ScholarOne Manuscripts

Author guidelines must be strictly followed. Please see:  Artificial Intelligence for a Sustainable Built Environment | Emerald Publishing

Authors should select (from the drop-down menu) the special issue title at the appropriate step in the submission process, i.e. in response to “Please select the issue you are submitting to”.

Submitted articles must not have been previously published, nor should they be under consideration for publication anywhere else while under review for this journal.

Key deadlines

Opening date for manuscripts submissions : 1st August 2026

Closing date for manuscripts submission: 31st December 2026

Closing date for abstract submission: 30th September 2026

Email for submissions: [email protected]