Gartner have recently reported that “80% of Today’s Project Management Tasks Will Be Eliminated by 2030 as Artificial Intelligence Takes Over”. Yet the penetration of AI within the construction industry has been slow. We believe that the key reasons for this are:
- A lack of understanding of advanced data analytics.
- A lack of quality data and consistency. Much of it exists in silos.
In 2018 SRM recognised the opportunity to leverage the exhaust plume of data that emerges from their projects to develop a bespoke planning tool to use the power of artificial intelligence (AI) as well as making the information more accessible for all team members.
Successful programming for a project is an art as much as a science. A detailed programme is developed by a project planner which is exported in various versions to different audiences. The accessibility of the data is often limited, and is quickly outdated. Queries are made in an ad hoc fashion using the planner to show a view or a list of tasks that a certain manager is interested in. This workflow creates a friction in accessing the information resulting in project teams often not carefully following the programme.
This creates significant project risks, impacting the ability to deliver projects on time, to budget and to a sufficient quality. Better engagement is required around the project programme, making it an accessible tool for all within the project – from project manager, to package managers, suppliers and subcontractors.
SRM hired Laminar Projects to create an improved programming application, making it more user friendly, visual and accessible, using live data. Using Microsoft PowerBI, we took the full extract of the dataset on a periodic basis, and with the data in its raw form, we created standard visualisations that satisfied questions that would be asked on an ongoing basis by various managers. The initial question set included:
- Which Critical Path Activities are happening now (broken down by zone/floor)?
- What activities are scheduled to complete in the next 2-6 weeks? Of those, which activities are behind schedule?
- What are the biggest risks and delays on site?
- On a plan, where should my contractor be floor-by-floor?
- What are the upcoming release dates for design/procurement?
Our new tool answers these questions by visualising the dataset in a user-friendly way. Typically the main audience for a construction programme is limited to the project manager, and the project planner. Our tool has shown that 14 people on the project, over half the team, have logged in and actively engaged with the programme, a significant improvement over a typical project.
As part of collaborative planning on the project, subcontractors are now being engaged on the project programme using the live interactive interface that has been developed. The subcontractors proposed construction programmes are integrated into the Master Strategic Plan, and then uploaded into the model. This then facilitates the communication of the critical path, procurement and supply chain risks, and package or subcontractor coordination requirements.
The 2 week look-ahead feature is specific to the package managers, allowing them for the first time to focus in on just their packages, and where they should sit within the programme. This allows for much closer collaboration with suppliers and their delivery against their programmes.
Already the solution is yielding real results, with a number of value-adding benefits to date including:
- A clear initiation of recovery plans by providing visibility of programme slip information.
- Retained focus on upcoming deliverables by providing consistent access to priority activities.
- Prevented on-site work area clashes by highlighting work areas visually.
- Programme data available to anyone on the project.
The renewed focus on which activities matter has led to reductions in on-site clashes and shortened programme timescales which leads to an indirect reduction in emissions where the overall construction lifecycle is concerned.
The next steps of this initiative can be summarised in two elements. The first is broadening the data being analysed and integrating the datasets together by mapping them in as many ways as possible.
The second step would be to roll out the analytics across the entire project portfolio to gather enough data to facilitate predictive analytics.
“It is showing the potential for what we can do with data, making it more visual and easier to interpret. It is also helping to improve the consistency in how we plan projects across the business, whilst reducing the reporting burden and increasing accuracy. As the model grows, AI and machine learning will mean we can more accurately predict risk on projects. This will ensure we deliver projects ahead of schedule, without costly and carbon intensive rework, and ensuring that we are thought leaders in changing how the industry operates.”Gareth Parkes, Company Knowledge Manager – Sir Robert McAlpine