Periodic Reporting for period 2 - COGITO (COnstruction-phase diGItal Twin mOdel)
Período documentado: 2021-11-01 hasta 2022-10-31
To minimise construction project duration/cost and reduce workplace accidents in construction, COGITO aims to contribute to the digitalisation of the construction industry by harmonising Digital Twins with BIMs and building a digital Construction 4.0 toolbox. This way, COGITO will contribute to productivity improvement and increased safety.
A bundle of services is developed and delivered within COGITO to facilitate the timely detection of health & safety hazards to humans, support timely identification of quality defects, and provide means for real-time workflow management on site. Methods to ensure interoperability among the different components comprising the COGITO architecture are applied within the Digital Twin Platform (DTP), the core of the entire toolchain. The COGITO Digital Twin relies on harmonised interfaces, standardised data structures, ontologies, communication protocols and data formats, delivering a re-usable and extensible construction digital twin.
The COGITO toolbox will be demonstrated at a pre-validation medium-scale infrastructure construction project and two validation construction sites to showcase its value. Beyond technical activities, extensive standardisation activities are carried out by COGITO partners to promote the COGITO standardisation proposals on construction, data modelling and linked open data to standardisation bodies, committees and working groups.
Furthermore, within COGITO, a wide range of the construction industry’s stakeholders are treated as research participants. A collaborative and engaging framework has been established, forming the COGITO Living Lab for exchanging opinions, experience, and knowledge. This framework facilitates the COGITO development based on agile principles and ensures the delivery of an innovative solution that meets actual stakeholders’ needs.
A bundle of services is developed and delivered within COGITO to facilitate the timely detection of health & safety hazards to humans, support timely identification of quality defects, and provide means for real-time workflow management on site. Methods to ensure interoperability among the different components comprising the COGITO architecture are applied within the Digital Twin Platform (DTP), the core of the entire toolchain. The COGITO Digital Twin relies on harmonised interfaces, standardised data structures, ontologies, communication protocols and data formats, delivering a re-usable and extensible construction digital twin.
The COGITO toolbox will be demonstrated at a pre-validation medium-scale infrastructure construction project and two validation construction sites to showcase its value. Beyond technical activities, extensive standardisation activities are carried out by COGITO partners to promote the COGITO standardisation proposals on construction, data modelling and linked open data to standardisation bodies, committees and working groups.
Furthermore, within COGITO, a wide range of the construction industry’s stakeholders are treated as research participants. A collaborative and engaging framework has been established, forming the COGITO Living Lab for exchanging opinions, experience, and knowledge. This framework facilitates the COGITO development based on agile principles and ensures the delivery of an innovative solution that meets actual stakeholders’ needs.
Within the first reporting period, management activities resulted in the delivery of two main documents, the Quality Assurance Plan, defining the management and operation framework of COGITO towards delivering high-quality results according to the Grant Agreement, and the Data Management Plan. The technical activities mainly focused on creating the project fundamentals. The main outcomes were the elicitation of the stakeholder requirements, the definition of the evaluation methodology; the regulatory and market analysis, and the detailed design of the system architecture. Furthermore, significant effort was made on identifying and analysing relevant existing data models and ontologies that led the development of the COGITO ontology network.
Within the period covered by this report, COGITO achieved most of the foreseen outcomes in line with the objectives of the project. Significant and sustained collaborative efforts have occurred. The results of those efforts have been described in 47 newly published deliverables. The main results achieved within this period include: (1) the dissemination, communication, exploitation and standardisation of the COGITO results’ planning; (2) the stakeholders’ participation in Living Lab, dissemination and communication activities, along with the preparation of material for their familiarisation with the COGITO tools; (3) the delivery of the second version of the COGITO architecture; (4) the development of all the components of the COGITO architecture; and (5) the delivery of the first version of the integrated COGITO ecosystem. In a nutshell, functional tools for Quality Control, Health & Safety, and Workflow Management, have been developed and documented, namely the IoT and Multisource Visual Data Pre-processing modules, the Digital Twin Platform, the Preventive, Proactive and Training Safety services, the Scan-vs-BIM Geometric and Deep Learning-based Visual control services, and the Adaptive Workflow Modelling and Management Automation tools, including the Blockchain-enabled Smart Contracts. Finally, as-planned (e.g. BIM Model, Schedules, Resources) as well as as-built data (e.g. point clouds, resources location data, images) have already been collected from a medium-scale infrastructure construction project, used for pre-validating tools in controlled conditions, following guidelines provided by the technical partners, and will constitute the main input for the pre-validation of the entire COGITO toolchain.
Within the period covered by this report, COGITO achieved most of the foreseen outcomes in line with the objectives of the project. Significant and sustained collaborative efforts have occurred. The results of those efforts have been described in 47 newly published deliverables. The main results achieved within this period include: (1) the dissemination, communication, exploitation and standardisation of the COGITO results’ planning; (2) the stakeholders’ participation in Living Lab, dissemination and communication activities, along with the preparation of material for their familiarisation with the COGITO tools; (3) the delivery of the second version of the COGITO architecture; (4) the development of all the components of the COGITO architecture; and (5) the delivery of the first version of the integrated COGITO ecosystem. In a nutshell, functional tools for Quality Control, Health & Safety, and Workflow Management, have been developed and documented, namely the IoT and Multisource Visual Data Pre-processing modules, the Digital Twin Platform, the Preventive, Proactive and Training Safety services, the Scan-vs-BIM Geometric and Deep Learning-based Visual control services, and the Adaptive Workflow Modelling and Management Automation tools, including the Blockchain-enabled Smart Contracts. Finally, as-planned (e.g. BIM Model, Schedules, Resources) as well as as-built data (e.g. point clouds, resources location data, images) have already been collected from a medium-scale infrastructure construction project, used for pre-validating tools in controlled conditions, following guidelines provided by the technical partners, and will constitute the main input for the pre-validation of the entire COGITO toolchain.
By the end of the project, key results include the delivery of an integrated COGITO solution, enabling the interaction of all its individual components and services, validated in actual construction projects, and the delivery of the COGITO data models and ontologies compliant with existing standards to the extend possible, to be promoted to standardisation bodies.
Progress beyond the state of the art is envisioned in a subset of the COGITO ecosystem. More specifically, in the DTP, aspects to be explored and advance the state of the art include design of a modular but fully extensible framework for change management, updating, consistency checking and version control of BIM data, linking contextual data with other information, information requirements and automated approaches for model checking, and automated generation of ‘digital’ twin models for the purposes of improvement at the construction site. Concering the H&S services, an existing digital twin application for H&S analysis will be extended and properly adapted to follow a more standardised approach, enriching the H&S checking prototype with more complex rules, and addressing expected computational runtime issues. To address these issues, a novel probabilistic Deductive Abductive Logic Programming framework to infer higher-level hypotheses of potential construction site hazards state based on likely, tentative explanations, will be developed. Furthermore, to detect rapidly unfolding risks in a dynamic construction site, up-to-date data are processed by state-of-the-art machine learning techniques, enabling the prediction of hazardous situations. In the context of the QC, a digital image processing pipeline for will be delivered, that goes beyond the current state of the art, mainly focusing on expanding the scope and effectiveness of automated QC to be able to detect structural geometric inconsistencies, cracks or other structural defects. For the Scan-vs-BIM Geometric QC, algorithms that implement existing Standard Test Methods for Geometric QC in a way that can automatically be applied to BIM objects to which reality capture data have been matched, will be developed, and the robustness of an existing Scan-vs-BIM solution in matching of as-build reality capture cloud points to as-design BIM elements, by innovatively integrating Scan-to-BIM techniques within the Scan-vs-BIM process, will be improved.
As a summary of quantified impacts, COGITO targets to the following KPIs: construction project cost savings, improved predictability of final project time/cost, increase of workforce productivity, and reduction in resource idle time by 20%; reduction in construction site accidents by 50%, promotion of COGITO activities to 5 standardisation bodies; at least one COGITO workshop with the attendance of standardisation body representatives to be conducted; at least 4 standards for which recommendations/punch-lists to be created; and finally, at least 4 Living Lab workshops to be conducted.
Progress beyond the state of the art is envisioned in a subset of the COGITO ecosystem. More specifically, in the DTP, aspects to be explored and advance the state of the art include design of a modular but fully extensible framework for change management, updating, consistency checking and version control of BIM data, linking contextual data with other information, information requirements and automated approaches for model checking, and automated generation of ‘digital’ twin models for the purposes of improvement at the construction site. Concering the H&S services, an existing digital twin application for H&S analysis will be extended and properly adapted to follow a more standardised approach, enriching the H&S checking prototype with more complex rules, and addressing expected computational runtime issues. To address these issues, a novel probabilistic Deductive Abductive Logic Programming framework to infer higher-level hypotheses of potential construction site hazards state based on likely, tentative explanations, will be developed. Furthermore, to detect rapidly unfolding risks in a dynamic construction site, up-to-date data are processed by state-of-the-art machine learning techniques, enabling the prediction of hazardous situations. In the context of the QC, a digital image processing pipeline for will be delivered, that goes beyond the current state of the art, mainly focusing on expanding the scope and effectiveness of automated QC to be able to detect structural geometric inconsistencies, cracks or other structural defects. For the Scan-vs-BIM Geometric QC, algorithms that implement existing Standard Test Methods for Geometric QC in a way that can automatically be applied to BIM objects to which reality capture data have been matched, will be developed, and the robustness of an existing Scan-vs-BIM solution in matching of as-build reality capture cloud points to as-design BIM elements, by innovatively integrating Scan-to-BIM techniques within the Scan-vs-BIM process, will be improved.
As a summary of quantified impacts, COGITO targets to the following KPIs: construction project cost savings, improved predictability of final project time/cost, increase of workforce productivity, and reduction in resource idle time by 20%; reduction in construction site accidents by 50%, promotion of COGITO activities to 5 standardisation bodies; at least one COGITO workshop with the attendance of standardisation body representatives to be conducted; at least 4 standards for which recommendations/punch-lists to be created; and finally, at least 4 Living Lab workshops to be conducted.