The Science of Anomalies: The Pillar of Predictive Maintenance in the DigiMaTRIA

The Challenge of Industrial Durability

In the Industry 5.0 landscape, the structural integrity of assets is a critical factor for operational continuity and economic sustainability. At the heart of this issue are sandwich panels — multi-layer construction systems widely adopted in industrial infrastructures due to their thermal efficiency and strength-to-weight ratio. However, exposure to harsh environments subjects these elements to complex degradation.

Within the scope of the DigiMaTRIA project, a deep understanding of anomalies is not merely a diagnostic exercise but the engine for creating an intelligent and predictive digital management system.

The Degradation Ecosystem: Agents and Mechanisms

Research developed by DigiMaTRIA identifies that the durability of sandwich panels is threatened by a synergy of external agents:

1. Chemical Agents and Corrosion: In coastal or industrial environments, the presence of chlorides, sulfur dioxide ($SO_2$), and high humidity accelerates galvanic and pitting corrosion. These agents attack protective steel layers, leading to section loss and compromising the adhesion between the faces and the core.

2. Physical and Mechanical Agents: Extreme thermal fluctuations cause differential expansion stresses, resulting in blistering phenomena or core delamination. Additionally, mechanical impacts and wind pressure impose loads that test the shear strength of the insulating material (PUR, PIR, or Mineral Wool).

3. Electromagnetic Agents (UV): Ultraviolet radiation degrades organic and polymeric coatings, making them brittle and allowing the infiltration of other aggressive agents.

4. Biological Agents: The development of fungi and microorganisms affects more than just aesthetics; biological activity can produce acidic byproducts that catalyze microbiologically influenced corrosion (MIC).

Typology of manipulation mechanisms by external agents, class and example.

From Anomalies to Data: The DigiMaTRIA Approach

The differentiator of the DigiMaTRIA project lies in the transposition of this physical-chemical knowledge into the digital domain. The detection of anomalies — such as discoloration, visible corrosion, deformation, or detachment — ceases to be a passive observation and becomes a data point in a BIM 7D model (Building Information Modeling focused on maintenance).

Project engineering utilizes Artificial Intelligence (AI) to classify these anomalies through computer vision, trained based on the degradation patterns identified in the literature review. By cross-referencing these images with IoT sensor data, DigiMaTRIA can predict the progression of a pathology before it reaches a critical state of failure.

Particle sizes of rust layers on the outer sheet exposed for 2 years to different types of atmospheres.

Life Cycle and Intelligent Maintenance

For engineers and asset managers, the importance of focusing on anomalies lies in Life Cycle Cost (LCC) optimization. Corrective maintenance is costly and disruptive; the predictive maintenance proposed by DigiMaTRIA is based on the "degradation signature" of each asset.

The integration of autonomous robotic agents for inspection allows for systematic and safe data collection, eliminating human error and ensuring that each detected anomaly is geo-referenced in the industrial unit's Digital Twin.

Blisters of different sizes on the outside façade of an industrial asset.

A New Frontier in Maintenance Engineering

Anomalies in sandwich panels are the symptoms of a system under stress. The DigiMaTRIA project elevates the analysis of these failures to a new technological level, where AI-assisted diagnosis and robotic management transform industrial asset preservation into an exact and efficient science.

Investing in DigiMaTRIA is an investment in the resilience of industrial infrastructures, ensuring that knowledge of material degradation translates into added value and long-term operational safety.

For more information, consult the scientific article presented at CIRMARE 2025, dedicated to the literature review of sandwich panel anomalies in industrial buildings.