Construction AIoT: Connecting Jobsite Data to Decisions
Construction sites are becoming increasingly connected.
A single project can involve job site equipment, workers, materials, access points, sensors, vehicles, inspections and more.
The value challenge is finding meaning in the disconnected data
We need to move from information about the site to information that helps people on the site.
This is where something called AIoT (Artificial Intelligence of Things) becomes interesting...
Let's explore what it means to connect physical construction resources to job site data.
What Does AIoT Really Mean In Construction?
This concept combines connected physical objects and elements with data processing, analytics, and AI capabilities to create value.
A very basic way to look at it is via this architectural diagram:
Physical Assets & Activities
↓
Sensors / Tags / Devices
↓
Connectivity
↓
Operational Data
↓
AI / Analytics
↓
Insights
↓
Human Decisions
The physical layer includes RFID tags, BLE location services, UWB positioning, GPS telematics, and similar tools. Each contributes unique signals that reflect different aspects of the job site.
All of these can be used to capture:
Information about physical objects (equipment, materials, workers, etc.)
Location of fixed and moving assets (equipment, vehicles, workers)
Job site logistics (material flow, equipment utilization, etc.)
Activities (construction, inspection, and more)
Importantly, the value stems not from disconnected data elements, but from the connections between them.
Let's Connect Some Dots
One easy example is to look at something as simple as an equipment tracking system.
In and of itself, it can show us where the tracked asset is at any given moment. A project management system, meanwhile, reflects the health of the overall project, and what activities are supposed to happen when.
Neither of these examples is particularly valuable on its own. But what happens if we connect these dots?
Something like equipment location + equipment activity + worker activity + project scheduling can be much more valuable than any of these elements considered separately.
This is the essence of what makes AIoT relevant to Y2M2's work. We are constantly looking for ways to connect the physical layer to create meaningful operational intelligence. Which leads us to our next point...
From Field Data To Operational Intelligence
Most, if not all, construction operations still revolve around spreadsheets, field service reports, job site inspections, and similar manual processes. This will not change simply because we will add connected devices and data processing.
But the data generated by these connected devices can make it easier to understand the state of the job site.
Manual reporting works great as long as we need information about known problems. But what can we do with the ability to continuously monitor the state of the project?
Equipment utilization is one excellent example of that. Another is the availability of materials and tools. The same goes for workers and access control data. Not to mention the potential to connect this information to project timelines and budgets.
More importantly, instead of asking "how much data can we get and from what sources?", we should be asking the question that really matters:
What Operational Question Can This Information Help Us Answer?
Understanding How Activities Align With Expectations
One area where the ability to combine different data sources would help dramatically is in understanding whether activities that happen match up with those that should happen (per the project timeline).
If we see signs of a particular type of activity at a certain location, we should be able to connect that to the project's expectations. Does that require anything specific to happen? Are the right people and materials present for it to happen?
Even if something does not match up, that still does not necessarily mean that it is a problem. We do not have to treat everything that happens on the job site as an issue. But we can take advantage of the anomaly detection capabilities offered by modern data science & big data processing tools.
When it comes to large-scale construction projects, this kind of approach is important simply because the sheer volume of data has to be processed efficiently. Instead of sifting via millions of individual data points, we can create relevant associations, detect exceptions, and use that to support a decision-making process.
Again, field operations and project management expertise matter as much as these technologies do. But having additional information that would be very difficult to obtain without it can be extremely valuable.
The Path To Better Construction Insights
Construction AIoT really does open up a lot of opportunities. But what specific steps should we take in actually building some of these capabilities?
A Y2M2 approach would be to design and implement stacks like this in a bottom-up manner:
We ask ourselves questions, such as:
- What operational question do we need answered?
- For which construction challenges do we want the added insight?
- What sorts of physical intelligence do we need?
What aspects do we need to analyze using this data?
Where are the relevant physical signals coming from?
- How can we bring this information together?
- What sorts of relationships exist between the data assets involved?
- What patterns, correlations, exceptions, and insights can we derive from it?
- How can we operationalize the conclusions for our operational benefit?
- What do we need to do to integrate it with our people, processes, and systems?
- How can we help stakeholders effectively utilize it?
This approach makes AI, machine learning, and big data merely enablers for a broader transformation.
The Bottom Line About Construction AIoT
At its core, it is about connecting the physical and digital dots on a construction site.
A platform such as the aforementioned CommCon AI is an excellent example of a broader move in this direction. This company puts its offerings under the "bringing construction workforce, equipment, materials, access and progress to a connected jobsite" banner.
The bottom line is that data is valuable in general. Connected data is significantly more valuable. And contextualized data has the potential to be even more so. And as construction continues to increase in complexity, this will become more and more important.
Top comments (0)