An innovative feasibility study into optimising construction site plant and equipment use by applying BIM, IoT and data analytics.
BIM Academy, in collaboration with BuildStream, Costain and Northumbria University have continued exploring how we can improve efficiency in utilisation of plant and equipment on site. This is achieved through the collation and aggregation of onsite data, combined with a construction programme and 3D model, allowing for a better understanding of current utilisation levels and productivity and more clearly identifying where improvements can be made both on live projects and future work.
Significant progress has been made in the third quarter of the project, allowing the team to progress into quarter four and present back final development in December. All required interviews and discussions with experts from Costain have been concluded, both in person and as a series of online workshops. Key data sets that were identified in quarter two were received, including legacy telematic data, current telematic data, building information models and construction programmes. The final interviews enabled the team to gain a thorough understanding of the data sets used to support a developed solution.
BuildStream have installed IoT devices on two units, capturing available data from the ECU (Electronic control unit) and CAN bus (Controller Area Network). Flexible intervals for the data are possible, allowing for the most appropriate data range / intervals to be selected. Flexible intervals for the delivery of data allows for more granularity and lends itself to a much more in depth understanding of plant and equipment performance and behaviour.
Costain have continued to provide a pilot site and assist in the organisation of interviews, discussions and access to data sets from experts within the company. A workshop was organised and attended by all team members to review progress and map out key interim dates and items needing completion to ensure milestones were met. This workshop proved to be immensely beneficial to the project and helped define the requirements for a project dashboard, identifying criteria for its success.
BIM Academy has continued to develop the 4D model, linking baseline and actual programmes to modelled objects to aid with the visualisation of project progress. Meetings on site with lead planners were arranged to run through the planning process, ensuring that a thorough understanding of how projects are planned and progress is monitored. Workshops were held with the wider team to understand information requirements and data formats essential to the project’s success.
Northumbria University has progressed work using historical data sets in parallel with BuildStream live data to create equipment operating performance predictions. These predictions will lead to better construction project decisions. Machine learning models have started to be developed, with legacy and BuildStream telematic data providing a rich vein of AI training data (in both variety and volume). This data has been used to develop the AI algorithms. A presentation of the research was delivered at the CIB W78 conference within the “Situation Awareness and Sensing Technologies” session. The presentation and work done to date was received well and a journal paper is currently in progress.
The team is now in the last stages of the research project, finalising individual strands of work and protocols for pulling all together in one single solution / platform. A presentation of the developed solution will be given to stakeholders in December, providing an opportunity for feedback to inform changes before the project deadline and final presentation back to Innovate UK.