Work package 2

Innovative Digital Solutions for Sustainable Deconstruction and Material Reuse 

Starting Month: 1/36

Lead Beneficiary : USTUTT

Objectives: WP2 aims to transform sustainable deconstruction and material reuse in construction through advanced digital technologies. The focus is on creating expert systems for accurate material characterization, enhancing traceability with RFID integration, establishing localized material banks for efficient reuse, and developing digital tools for informed deconstruction decision-making. These initiatives exploit AI, ML, AR, blockchain, and IoT to promote CE principles, optimize resource utilization, and ensure environmental sustainability in construction practices.

Description of Work: 

Task 2.1. Development of expert systems for material characterization and sorting (Leader: NTUA, PM=20.0M): This task involves the deployment of specific AI and ML models such as CNNs for image recognition of materials and RNNs for analysing time series data from sorting processes. Decision trees and clustering algorithms will be employed to categorize materials based on various attributes like composition, condition, and recyclability. Staff involved in this task are chosen for their deep expertise in AI, ML, AR, blockchain, and IoT technologies, ensuring a multidisciplinary approach to developing sophisticated expert systems for material characterization and sorting. Role of Participants: NTUA: Leads with AI and ML model development, utilizing CNNs for material image recognition and RNNs for process analysis. PRISMA: Implements IoT technologies and integrates sensors. AIDIGITS: Ensures data integrity and traceability through blockchain-based smart contracts. BUILTCOLAB: Employs AR technologies, specifically Marker based AR and 3D point cloud models, to support manual sorting operations with real-time data overlays. UOA: Uses digital twin simulations to optimize sorting processes, applying advanced virtual models for environment testing. 

Task 2.2. Integration of RFID technology to enhance material traceability (Leader: PRISMA, PM=14.0M): The task will utilize RFID data models and middleware to efficiently capture, process, and manage the vast amount of data generated by RFID tags and readers. This includes employing Entity-Relationship (ER) models for organizing RFID-related data and Complex Event Processing (CEP) models for real-time analysis of RFID data streams, enabling immediate actions based on material location and status. Role of Participants: PRISMA: Spearheads the task with RFID data models and middleware implementation, ensuring efficient data capture and management. NAPSLAB: Incorporates predictive analytics and digital twin models to forecast material usage and lifecycle stages from RFID data. NTUA: Integrates RFID tracking data with structural materials databases, enhancing material traceability in construction projects. 

Task 2.3. Establishment of localized material banks for effective reuse (Leader: USTUTT, PM=14.0M): This involves the development and application of GIS models to map available materials and their locations, facilitating efficient material logistics and reuse. Decision support models, such as Multi-Criteria Decision Analysis (MCDA), will be employed to evaluate and prioritize materials for reuse based on various parameters like quality, environmental impact, and proximity. Role of Participants: USTUTT: Leads the establishment of material banks, using GIS models to map materials and MCDA for material evaluation and prioritization. UOA: Assesses the compatibility of bank materials for AM and integrates them into BIM. INS: Incorporates LCA models to quantify the environmental benefits of material reuse, reinforcing the project's commitment to sustainability

Task 2.4. Innovative digital tools for decision-making in deconstruction processes (Leader: AIDIGITS, PM=18.0M): This involves leveraging advanced simulation models to predict the outcomes of various deconstruction scenarios, enabling stakeholders to make informed decisions. These tools incorporate elements of ML algorithms, to analyze historical deconstruction data and predict the most efficient methods and processes for material recovery considering factors such as material type, building architecture, and environmental impact. Dynamic virtual replicas of buildings slated for deconstruction will be developed, allowing for the simulation and assessment of various deconstruction approaches in a virtual environment. Role of Participants: NTUA: Spearheads the development of digital decision-making tools, integrating simulation models and AI to forecast deconstruction outcomes. JUST: Enhances the task with AI-driven predictive models that evaluate deconstruction strategies, ensuring optimal material recovery. AIDIGITS: Implements digital twin technology to simulate deconstruction processes, providing a virtual testing ground for decision-making.