Information is critical to the successful implementation of technologies like AI and machine learning, yet all too often infrastructure project data is not being reused. Claire Rutkowski of Bentley Systems takes a look
Our world runs on data. The advent of BIM and the proliferation of IoT-sensing devices have driven a veritable deluge of data. We are drowning in it. But surprisingly, we are not using it.
A report by FMI Corporation, a consultancy firm, stated that 96% of infrastructure project data specifically is never reused.
You could argue that a certain portion of that data should not be reused because it is project-specific, but certainly not 96% of it.
One of the reasons we are unable to leverage infrastructure project data is that every time we hand a deliverable off from one firm to another or from one phase of the asset lifecycle to another (e.g. the handoff from project design and delivery to construction or construction to commissioning), we lose data.
90% of the total data is unstructured
Perhaps the tools or systems being used in each phase do not talk to one another, or the data being generated in one phase is incompatible with the structure of the data in the next phase.
As a result, we lose a lot of the hard work performed in one phase and in many ways start over in the next phase – with less fidelity and detail.
The same FMI study reported that 90% of the total data its accounts produce during engineering and construction is unstructured.
Unstructured data can be difficult, if not impossible, to translate and migrate from one tool to another; it is one of the main drivers for information getting lost along the way.
The result is a lack of visibility into data, leading to challenges in communication and collaboration because stakeholders are not looking at or talking about the same datasets.
It also means that decision-makers are making decisions based on incomplete data, because some of it did not transfer over. And siloed data gets left behind, driving obsolescence.
Leveraging AI and machine learning to deliver sustainable outcomes
For the infrastructure sector to take full advantage of tools like artificial intelligence and machine learning, stakeholders need robust datasets containing structured and unstructured data that can be mined for patterns and insights.
AEC firms and asset operators need cohesive, complete datasets to foster communication and collaboration.
Cohesive and complete data sets can also ensure that we are measuring the complete carbon footprint of an asset, from project delivery to construction to operations and potential decommissioning.
This will help the sector drive sustainable outcomes.
And robust and accurate data that can be leveraged by machine learning and artificial intelligence tools will also help ensure accuracy to design intent, greater efficiencies throughout the asset lifecycle, and provide insights not otherwise gleaned.
How can the infrastructure sector achieve this?
From a technology perspective, the creation and maintenance of rich datasets require open platforms that can work with multiple tools.
Interoperability among those tools is also required, and the data must be able to be shared securely and appropriately throughout the entire supply chain.
The data also needs to be captured and maintained. Data requirements must be standard across the entire lifecycle.
We use P&IDs to tie things together, but let’s go deeper.
If we move data definitions further up the lifecycle and define what is needed through each phase, we can ensure that the right data is captured as projects progress.
Harnessing digital twins capabilities
By using digital twin capabilities, infrastructure stakeholders can combine and amalgamate data from various sources, making a series of data siloes into one cohesive bank of geospatially referenced data points about an asset, whether that is a carbon footprint, emissions, as-designed data, as-built data, maintenance information or asset performance information.
However, even digital twins can become siloed and stagnant if not done right.
To truly be valuable across the infrastructure lifecycle, digital twin technologies need to have an open foundation to build upon with infrastructure schemas that have standards for data translation from one application to another.
Only then can a digital twin be leveraged as a dynamic, comprehensive living dataset.
Only then can we light up all the dark data we currently lose throughout the infrastructure and asset lifecycle to gain true insights.
Is the work worth the effort? Absolutely.
Rich datasets drive actionable insights to improve project decision-making and operational performance.
These insights can shed light on previously undetected trends, allowing an organisation to proactively course correct, leading to better outcomes for all stakeholders – design firms, EPCs, constructors, asset owners and operators, and the public at large.
It’s all about the data
Stakeholders can then leverage machine learning, artificial intelligence and other tools, which thrive on rich data to glean insights into their data for carbon calculations, predictive maintenance recommendations and, better, more sustainable outcomes.
But it all comes back to the data.
Let us not settle for 4% data reuse. The opportunity for smarter infrastructure outcomes across the lifecycle is in the 96%.
By using infrastructure project data as a foundation, we can collaborate, leverage, reuse and become much more efficient and insightful.
Given the increased demand for infrastructure projects, the growing backlogs, resource shortages and the urgent need to reduce climate change, we must get the fundamentals right. We cannot afford not to.
Claire Rutkowski
SVP and CIO champion
Bentley Systems
