Periodic Reporting for period 1 - ACCORD (Automated Compliance Checks for Construction, Renovation or Demolition Works)
Période du rapport: 2022-09-01 au 2024-02-29
Building regulations, standards, and requirements are essential in designing and constructing safe, high-quality buildings. The official compliance process necessitates extensive documentation across multiple formats, such as documents, drawings, calculation sheets, models, and simulations. Better information exchange management and automatic code-compliance checking would reduce construction costs and improve built quality.
ACCORD's objective is to digitalise building permit and compliance processes using BIM and other data sources to improve the productivity and quality of design and construction processes, support the design of climate-neutral buildings, and advance a sustainable built environment in line with the EU Green Deal and new European Bauhaus initiative. These digitized processes must be human-centred, transparent, and cost-effective for the permit applicants and authorities. Another objective is to develop and integrate technical solutions for automating compliance checking of buildings in their design, construction, and renovation/demolition lifecycle phases. All this will be based on open and neutral data exchange standards.
ACCORD supports the achievement of the EU Green Deal by 1) developing open API specifications allowing local authorities to choose their digital services without a lock-in, 2) automating the building compliance checking, 3) creating information guidelines for structured data models to increase the value of buildings and reduce their operating costs and 4) providing an open semantic framework based on microservices allowing businesses to connect and develop solutions resulting in a scalable, durable and flexible permitting ecosystem.
ACCORD's objective is to digitalise building permit and compliance processes using BIM and other data sources to improve the productivity and quality of design and construction processes, support the design of climate-neutral buildings, and advance a sustainable built environment in line with the EU Green Deal and new European Bauhaus initiative. These digitized processes must be human-centred, transparent, and cost-effective for the permit applicants and authorities. Another objective is to develop and integrate technical solutions for automating compliance checking of buildings in their design, construction, and renovation/demolition lifecycle phases. All this will be based on open and neutral data exchange standards.
ACCORD supports the achievement of the EU Green Deal by 1) developing open API specifications allowing local authorities to choose their digital services without a lock-in, 2) automating the building compliance checking, 3) creating information guidelines for structured data models to increase the value of buildings and reduce their operating costs and 4) providing an open semantic framework based on microservices allowing businesses to connect and develop solutions resulting in a scalable, durable and flexible permitting ecosystem.
ACCORD analysed the complex landscape of built environment compliance checking and permitting across Europe to ascertain the user requirements for future digitalization. The current building permit processes of the demonstration countries (Finland, Estonia, the UK, Germany and Spain) were modelled.
ACCORD developed a semantic framework for European building permit processes, regulations, data and tools. This framework will drive rule formalization and integration of existing compliance tools as microservices. ACCORD elicited the technical requirements for developing and integrating micro-services into the ACCORD project.
ACCORD defined a cloud-architecture and its components, which will be developed and utilized in the project. The components will provide consistency, interoperability and reliability with national regulatory frameworks, processes and standards. The components will be implemented and demonstrated across construction projects in various EU regulatory contexts.
ACCORD developed a methodology for developing machine-readable rules from selected country-specific building regulations. It includes a domain-specific language that ACCORD develops. Two approaches will be used. The first is manual, using the RASE (Requirement, Applicability, Exception, Selection). The second is AI-assisted, applying natural language processing, which has produced an annotated CODE-ACCORD dataset comprising 864 finely curated sentences extracted from English and Finnish building regulations, focusing on self-contained rules expressed therein. The dataset is a significant asset in developing intelligent systems and tools that can contribute to automated regulatory compliance and efficiency in the construction industry. It can be used to apply different Machine Learning and Deep Learning related tasks, such as classification tasks.
ACCORD created an AEC Compliance Checking and Permitting Ontology (AEC3PO) to provide a set of formalized semantics to underpin this approach. ACCORD has also reviewed existing ontologies, rule languages, standards, and data models within the real estate and construction domain to determine their suitability for automated compliance checks.
ACCORD developed a semantic framework for European building permit processes, regulations, data and tools. This framework will drive rule formalization and integration of existing compliance tools as microservices. ACCORD elicited the technical requirements for developing and integrating micro-services into the ACCORD project.
ACCORD defined a cloud-architecture and its components, which will be developed and utilized in the project. The components will provide consistency, interoperability and reliability with national regulatory frameworks, processes and standards. The components will be implemented and demonstrated across construction projects in various EU regulatory contexts.
ACCORD developed a methodology for developing machine-readable rules from selected country-specific building regulations. It includes a domain-specific language that ACCORD develops. Two approaches will be used. The first is manual, using the RASE (Requirement, Applicability, Exception, Selection). The second is AI-assisted, applying natural language processing, which has produced an annotated CODE-ACCORD dataset comprising 864 finely curated sentences extracted from English and Finnish building regulations, focusing on self-contained rules expressed therein. The dataset is a significant asset in developing intelligent systems and tools that can contribute to automated regulatory compliance and efficiency in the construction industry. It can be used to apply different Machine Learning and Deep Learning related tasks, such as classification tasks.
ACCORD created an AEC Compliance Checking and Permitting Ontology (AEC3PO) to provide a set of formalized semantics to underpin this approach. ACCORD has also reviewed existing ontologies, rule languages, standards, and data models within the real estate and construction domain to determine their suitability for automated compliance checks.
ACCORD’s key exploitable results can be grouped into the following thematical groups: A) Methods and tools for formalising documents into machine-readable formats, B) Technical frameworks, standards and harmonised guidelines for software and platform development for the use-case of automated digital building permit process, C) Software components for building permitting and compliance checking, D) Guidelines and information requirements for BIM-based building permit process, and E) Datasets.
ACCORD has developed a semantic framework for automatic compliance checking and cloud architecture, specifying the role of each component and interoperation between the components. The framework supports the business development around compliance checking as it specifies the market for different interoperable software. This framework advances the state of the art by providing the first distributed digital building permitting framework that is underpinned by formalized semantics.
ACCORD has developed AEC3PO (Architecture, Engineering and Construction Compliance Checking Ontology). This is the first construction-focused compliance checking ontology, which allows the modelling of a) building codes, regulations, and standards, b) integration of rules authored using various methods, c) compliance and building permit processes and documentation, and d) compliance and permitting actors. Together with a rule format, this ontology allows the building of a new industry that interprets and converts building codes to machine-readable formats.
ACCORD developed a regulation digitisation methodology. This methodology has advanced the state of the art as it integrates established regulatory compliance methods (the RASE approach) that can be formalized in a semantic domain-specific language based on the AEC3PO ontology. The methodology is further innovative as it formalizes the integration of two approaches: AI-assisted, applying natural language processing, and manual. In the future, these two approaches will be integrated further to produce the first hybrid automated compliance checking methodology.
ACCORD produced CODE-ACCORD, a dataset comprising 864 finely curated sentences extracted from English and Finnish building regulations, focusing on self-contained rules expressed therein. The dataset is a significant asset in fostering the development of intelligent systems and tools that can contribute to automated regulatory compliance and efficiency in the construction industry. It can be used to apply different Machine Learning and Deep Learning tasks, such as classification tasks.
ACCORD has developed a semantic framework for automatic compliance checking and cloud architecture, specifying the role of each component and interoperation between the components. The framework supports the business development around compliance checking as it specifies the market for different interoperable software. This framework advances the state of the art by providing the first distributed digital building permitting framework that is underpinned by formalized semantics.
ACCORD has developed AEC3PO (Architecture, Engineering and Construction Compliance Checking Ontology). This is the first construction-focused compliance checking ontology, which allows the modelling of a) building codes, regulations, and standards, b) integration of rules authored using various methods, c) compliance and building permit processes and documentation, and d) compliance and permitting actors. Together with a rule format, this ontology allows the building of a new industry that interprets and converts building codes to machine-readable formats.
ACCORD developed a regulation digitisation methodology. This methodology has advanced the state of the art as it integrates established regulatory compliance methods (the RASE approach) that can be formalized in a semantic domain-specific language based on the AEC3PO ontology. The methodology is further innovative as it formalizes the integration of two approaches: AI-assisted, applying natural language processing, and manual. In the future, these two approaches will be integrated further to produce the first hybrid automated compliance checking methodology.
ACCORD produced CODE-ACCORD, a dataset comprising 864 finely curated sentences extracted from English and Finnish building regulations, focusing on self-contained rules expressed therein. The dataset is a significant asset in fostering the development of intelligent systems and tools that can contribute to automated regulatory compliance and efficiency in the construction industry. It can be used to apply different Machine Learning and Deep Learning tasks, such as classification tasks.