Ornella Iuorio

The stone industry generates substantial waste along the quarrying, cutting and finishing chain, with only about 22% of the extracted material processed into final products, while roughly 64% is lost as slurry and 8% as solid waste. Compared to other industries such as wood and steel, there is a significant gap in the exploitation of relevant circular economy indicators for the stone industry. To address such limitation, this study proposes a sector-specific indicator framework built in three stages: systematic identification of existing CE indicators through a Scopus-based literature review, further refinement against three strategies for the stone sector (materials preservation, process optimisation and reduction of risk to human well-being), together with two suitability filters based on development stage and redundancy, and finally an indicative prioritisation using a simplified Analytic Hierarchy Process (AHP) procedure implemented in Super Decisions. Seven indicators are retained as the most applicable to the sector: Energy Use, Water Consumption, Greenhouse Gas Emissions, Replacement Ratio, Generation Ratio, Upcycling Score and Product Quality Index. The Consistency Ratio is 0.070 for the criteria pairwise-comparison matrix and ranges from 0.000 to 0.010 across the six indicator pairwise-comparison matrices, reflecting the internal coherence of the deterministic scoring procedure used to derive them. The framework is presented as an initial, transparent proposal to support monitoring and resource efficiency benchmarking in the stone supply chain. Empirical validation through further industrial case studies and broader expert contribution is an essential next step.

Smart building management and HVAC control increasingly relies on granular, real-time occupancy data to bridge the building performance gap. However, traditional monitoring methods, such as camera-based systems, are often limited by high costs and significant privacy concerns. This study proposes an alternative ”virtual sensor” methodology that leverages Artificial Neural Networks (ANN) to estimate occupancy counts using existing Indoor Environmental Quality (IEQ) readings (CO2, temperature, humidity). To overcome the common challenge of model generalisability, dynamic time-series sensor data are fused with static spatial features (e.g., room area) extracted from a Building Information Model (BIM). The ANN architecture is trained and validated against ground-truth occupancy counts from a privacy-masked camera system across 20 diverse building spaces. A feature selection process utilising Mutual Information and hierarchical clustering identifies rolling temporal statistics as superior predictors compared to raw sensor data. The results demonstrate that the full ANN model achieves high fidelity with a Mean Absolute Error (MAE) of 0.65, significantly outperforming naïve baselines. The integration of BIM spatial data allows the model to adapt to varying room configurations without extensive re-training. Because minimum ventilation rates under EN 16798-1 scale with occupant count, a reliable, transferable occupancy stream is the enabling input for occupancy-responsive operation such as demand-controlled ventilation, supplied here from existing IEQ sensors, without the cost or privacy burden of camera-based counting. With the described methodology, this research provides a scalable, privacy-preserving foundation for enhancing intelligent building automation, optimising energy consumption, and improving occupant well-being through data-rich Digital Twins.

Accurate state estimation of building components across large workspaces is crucial for automated construction assembly, yet construction environments are inherently indeterminate due to material variability, structural deflection and changing site conditions. This paper addresses the resulting tension between precision and adaptability by presenting a dual-camera and point-cloud fusion framework for robotic assembly of timber panels. A tripod-mounted global RGB-D camera provides continuous workspace coverage, while an eye-in-hand camera supplies high-precision local measurements when components enter its field of view. AprilTag fiducial markers on panels and the robot base support continuous self-calibration of the global camera relative to the robot, and statistical fusion combines pose estimates through distance-based confidence weighting, inverse-variance weighting and temporal outlier rejection. These fused poses are optionally refined using point cloud registration via a coarse-to-fine Iterative Closest Point (ICP) scheme. The system is embedded in a ROS-based architecture that links real-time sensing to a parametric design environment and an impedance-controlled Kuka iiwa manipulator. Experiments on the insertion of interlocking timber panels show that dual-camera fusion substantially improves consistency over global-only sensing and enables successful assembly where manual calibration or single-sensor feedback fails. The results demonstrate how balancing precision sensing with tolerance to uncertainty can support robust, adaptive robotic construction workflows.

This paper presents a novel framework for estimating urban tree dimensions and their associated CO2 sequestration potential using neural network-based image segmentation applied to Google Street View imagery. Our approach combines a dual image acquisition strategy with geometric modelling based on the pinhole camera lens model to automatically extract key tree metrics such as trunk diameter and overall height. By leveraging road width data to approximate the distance to trees and applying trigonometric equations, pixel measurements are converted into real-world dimensions. To facilitate large-scale analysis, a custom neural network that builds on established models was developed – fine-tuning a Faster R-CNN/ResNet-50-FPN architecture with extensive data augmentation from both the Urban Street Tree Dataset and local imagery. Validation against ground truth measurements shows that while tree height estimates achieve high accuracy (within 3% error), diameter estimations exhibit greater variability, highlighting areas for future improvement. The derived tree dimensions are then used to compute CO2 sequestration using standard forestry equations, demonstrating the method’s potential to inform urban sustainability assessments. Overall, this integrated pipeline offers a scalable solution for urban tree inventory and environmental impact studies.

The increasing frequency and intensity of climate-related and seismic events globally call for innovative, resilient approaches to social housing development. This study explores the potential of bamboo, as an abundant, locally available, rapidly renewable, and structurally resilient biobased material, to serve as the foundation for climate-adapting neighbourhood’s creation in social housing resettlements in the coastal region of Manabí, Ecuador. By adopting a cross-sectoral collaboration approach, and integrating insights from architecture, materials science, structural engineering, and human geography, a multi-domain approach is proposed. This was defined through the engagement with a broad range of stakeholders, including local architects, international firms, local risk managers and NGOs. The proposed approach demonstrates that bamboo is a practical and adaptable solution for housing projects creation or adaptation, not only enhancing structural resilience, but also reinforcing social cohesion and local empowerment in vulnerable Latin American neighbourhoods. The utilisation of bamboo also addresses simultaneously environmental, economic, and social challenges, fostering the growth of communities, which are most affected by such hazards, while respecting ecological sustainability goals. The findings reported here are the result of an international interdisciplinary research programme aiming to define strategies for development of housing and neighbourhood for wellbeing, so that they are structurally resilient, low-carbon, and allow sustainable growth of communities affected by multiple hazards in Latin America.

Robotic assembly of shell systems requires rethinking about how shell components are connected together. This research proposes a new standardized connection system for single curvature panelised systems. A triad connection at the vertices of a hexagonally tessellated structure has three independent variables, representing the three angles of the adjacent panels. The combinations of the different variables in a single shell structure produces hundreds of solutions where each solution is an independent connection. The connection system proposed solves this problem by having three articulating connecting fingers around a central hub. This connection system relies on a parametric environment and adapts to a variety of shell geometries and curvatures. The parameters of the design dictate the joint position and only reflect on the design of the panels. This type of standardized connection offers flexibility in the design of the structure, as well as being suitable for repurposing and use in other projects. The connections are designed for 3D printing “in-place” to reduce assembly and post-processing. The nature of this design makes it inherently easy to adjust for robotic assembly by changing the central hub features for easier manipulation by a robotic gripper. The validity of this solution is assessed in this work through tensile testing.