Subventions et des contributions :

Subvention ou bourse octroyée s'appliquant à plus d'un exercice financier. (2017-2018 à 2022-2023)

The identification, modeling, and evaluation of design & construction processes have recently attracted attention within the Architecture, Engineering, Construction & Owners (AECO) body of research. Building Information Modeling (BIM) supports project teams through such processes and therefore can be effectively used as an Enterprise Resource Planning (ERP) tool for AECO industry. BIM can integrate distributed information and derive new information. However, providing actionable information requires identification and evaluation of the end-to-end processes, as they happen in practice (opposed to what has been planned).
Process mining, a branch of data science, detects end-to-end business processes via pattern recognition in enterprise operation data. For this purpose, data from ERP systems is processed to form event-logs (inputs of process mining engines). In spite of substantial contributions to enhancing processes in a wide range of industries; process mining has not had any major applications, in design & construction processes. This is primarily due to the lack of ERP systems in AECO, and furthermore the need for domain-specific solutions to reflect unique characteristics of design & construction projects.

The overarching goal of this program is to automate identification and evaluation of design & construction processes and the associated actor networks. This goal will be achieved through three main objectives: (1) Extending the existing data standards for supporting BIM to generate design & construction event-logs; (2) Identifying, evaluating and enhancing design & construction processes via analysis of dynamic IFC files; (3) Providing decision support in form of practice-oriented KPIs and enhanced simulation models. Seven HQP will be trained with dual specialties (design & construction knowledge; and data science skills) who will contribute to achieving these objectives.

This program will extend the existing open standards of event-logs to support events, attributes and relationships in design & construction processes. The standard will then be used to create a service to generate BIM event-logs from IFC files. In the second phase, a variety of algorithms will be created to capture some of the generic design & construction processes. Using those algorithms, models will be trained for performance of those processes (e.g., deviations, bottlenecks, etc.); as well as the behavior of associated actors. Finally, the program will offer a set of practice-oriented KPIs and a method to enhance construction process simulation.

This research program will be a major step forward in assessment of processes and actors in AECO. It will yield deliverables at three major levels: (1) data standard and data mining tools (2) methods and algorithms (3) models & KPIs. These will particularly benefit three major domains: academia; global and Canadian AECO industry; and public sector bodies.