PSI - Issue 84

Regina Finocchiaro et al. / Procedia Structural Integrity 84 (2026) 1264–1269

1265

Keywords: existing bridges; visual inspections; risk management; Gestalt; human factors; data quality; decision support.

1. Introduction of the regulatory framework and multi-level procedure of the Bridge Guidelines In recent years, the existing bridges safety has become a structural issue, both in technical sense and political sense, for the management of road networks (you have to take into account that according to the Italian Code a bridge is a six-meters span structure: for this reason, bridge number is very high). The approach introduced by National Guidelines of Infrastructure Ministry seeks to move beyond a purely “failure-driven” logic and to establish a replicable procedure based on inventory surveys, visual inspections, risk classification and, if necessary, safety verification and monitoring. Despite a robust methodological architecture, a decisive part of the information chain remains dependent on the perception and judgment of the observer: visual inspection must recognize and describe defects, deterioration and anomalies; it must group them into coherent phenomena (e.g., a cracking pattern or a carbonation and cover delamination area for r.c. structures; fatigue crack, corrosion and bolt loosening for steel structures, a recurring set of defects in bearings) and must distinguish structurally meaningful signals from “noise” (superficial stains, chromatic inhomogeneities, traces of previous maintenance). From a cognitive standpoint, an inspector must perform an organization of the perceptual field. Gestalt psychology provides a vocabulary and a set of principles to describe how humans organize the visual stimuli into meaningful forms. These principles are already active during any on-site survey: proximity leads to grouping nearby defects; similarity supports the perception and identification of “similar problems”, aggregating, for instance, corrosion events having comparable morphology; good continuation leads, for example, one to follow a crack as a single phenomenon; closure induces a mental reconstruction of missing contours; figure–ground and salience guide attention toward what stands out from the context. The point is not to turn applied engineering into applied psychology. The statement appears simple, but it has uncomfortable risks: if the technical procedure systematically depends on a perceptual process, then it is necessary to make this process an explicit, controllable and trainable process. This paper proposes an integration of Gestalt-based perceptual mechanisms into the Bridge Guidelines as a means to reduce discretion and increase data quality, without introducing new Code prescriptions but by improving operational implementation. Paper main objectives are: 1. to discuss where and how Gestalt principles implicitly enter in the multi-level procedure 2. to propose a set of “perceptual rules” translatable into operational indications for L0  L3/L4 3. to define perceptual quality indicators integrable into standard inspection forms and information systems 4. to demonstrate the approach effect on Class of Attention determination and further investigations planning. The Authors intend to enlarge this procedure [considering salience, ambiguity, coherence and knowingness (SACK method)] through illustrative examples and math formulas that will be based on present paper proposal. 2. Regulatory framework and multi-level procedure of the Bridge Guidelines The Italian Guidelines for existing bridges are implemented through a framework linking inventory surveys, inspections and safety evaluations within a multi-objective pathway. The 2022, July 1 Ministry Decree adopted actual Guidelines, replacing previous Annex and extending their application in a national framework and mandatory scheduling. The approach is structured at different levels: a rapid assessment at network scale (inventory, inspections, and classification) and a more targeted and complex evaluation for the single asset; those levels are: • Level 0: structured inventory survey; collection of geometric, typological, functional, and contextual data. • Level 1: visual inspections and compilation of defect/deterioration standard sheets, with verification and integration of the inventory data. • Level 2: estimation of single Class of Attention (CdA) for different risk types with a final combination into an overall CdA, according to a guided procedure. 1.1. Paper objectives