Purpose Buildings and their use is a complex process from design to occupation. Buildings produce huge volumes of data such as building information modelling (BIM), sensor (e.g. from building management systems), occupant and building maintenance data. These data can be spread across multiple disconnected systems in numerous formats, making their combined analysis difficult. The purpose of this paper is to bring these sources of data together, to provide a more complete account of a building and, consequently, a more comprehensive basis for understanding and managing its performance. Design/methodology/approach Building data from a sample of newly constructed housing units were analysed, several properties were identified for the study and sensors deployed. A sensor agnostic platform for visualising real-time building performance data was developed. Findings Data sources from both sensor data and qualitative questionnaire were analysed and a matrix of elements affecting building performance in areas such as energy use, comfort use, integration with technology was presented. In addition, a prototype sensor visualisation platform was designed to connect in-use performance data to BIM. Originality/value This work presents initial findings from a post occupancy evaluation utilising sensor data. The work attempts to address the issues of BIM in-use scenarios for housing sector. A prototype was developed which can be fully developed and replicated to wider housing projects. The findings can better address how indoor thermal comfort parameters can be used to improve housing stock and even address elements such as machine learning for better buildings.

No abstract

The operation and maintenance of built assets is crucial for optimising their whole life cost and efficiency. Historically, however, there has been a general failure in the transfer information between the design-and-construct (D&C) and operate-and-maintain (O&M) phases of the asset lifecycle. The recent steady uptake of digital technologies, such as Building Information Modelling (BIM) in the D&C phase has been accompanied by an expectation that this would enable better transfer of information to those responsible for O&M. Progress has been slow, with practitioners being unsure as to how to incorporate BIM into their working practices. Three types of challenge are identified, related to communication, experience and technology. In examining the last aspect, it appears that a major problem has been that of interoperability between building information models and the many computer-aided facilities management (CAFM) systems in use. The successful and automatic transfer of information from a building model to an FM tool is, in theory, achievable through the medium of the Industry Foundation Classes (IFC) schema. However, this relies upon the authoring of the model in terms of how well its structure permits the identification of relevant objects, their relationships and attributes. The testing of over 100 anonymised building models revealed that very few did; prohibiting their straightforward mapping to the maintenance database we had selected for the test. An alternative, hybrid approach was developed using an open-source software toolkit to identify objects by their geometry as well as their classification, thus enabling their automatic transfer. In some cases, manual transfer proved necessary. The implications are that while these problems can be overcome on a case-by-case basis, interoperability between D&C and O&M systems will not become standard until it is accommodated by appropriate and informed authoring of building models.

Heavy equipment represents a major cost element and a critical resource in large infrastructure projects. Automating the measurement of their productivity is important to remove the inaccuracies and inefficiencies of current manual measurement processes and to improve the performance of projects. Existing studies have prevalently focused on equipment activity recognition using mainly vision based systems which require intrusive field installation and the application of more computationally demanding methods. This study aims to automate the measurement of equipment productivity using a combination of smartphone sensors to collect kinematic and noise data and deep learning algorithms. Different combination inputs and deep learning methods were implemented and tested in a real-world case study of a demolition activity. The results demonstrated very high accuracy (99.78%) in measuring the productivity of the excavator. Construction projects can benefit from the proposed method to automate productivity measurement, identify equipment inefficiencies in near real-time, and inform corrective actions.

Site equipment represent a major cost element in construction projects. Measuring equipment productivity help to identify equipment inefficiencies and improve their productivity; however, measurement processes are time and resource intensive. Current literature has focused on automating equipment activity capture but still lack adequate approaches for measurement of equipment productivity rates. Our contribution is to present a methodology for automating equipment productivity measurement using kinematic and noise data collected through smartphone sensors from within equipment and deep learning algorithms for recognizing equipment states. The testing of the proposed method in a real world case study demonstrated very high accuracy of 99.78% in measuring productivity of an excavator.