Commercial construction sites of the future are being connected.
Equipment can generate telemetry, RFID and BLE tags can identify assets or materials, UWB can offer higher accuracy location data, GPS can help track machines, and sensors may observe temperature, vibrations, or machine state.
But the question of what to do with this data is only partially answered.
The more interesting engineering problem is what to do with this data after it is created.
How to get this information from the physical world into software, how to process it, how to relate it to other data in the ecosystem, and how to use it for good?
That’s where the idea of AIoT (AI + IoT) helps to build an interesting architecture for the commercial construction industry.
From Things to Information
A connected jobsite environment can be imagined as a chain:
Physical Assets → Sensors → Connectivity → Data → Processing → AI/Analytics → Decision → Action
Each link in this chain serves a particular purpose.
Sensors, tags, and RFID readers are used to read information from the physical world.
Connectivity passes this information through networks.
Gateways and edge systems can process this information locally.
Cloud or enterprise data platforms can be used to store this information for later analysis or combine it with other data sources.
AI adds value when it examines patterns, connections, or outliers in large data sets.
This is where the value of AIoT lies not in isolated sensors attached to construction equipment.
The bigger picture shows that there is value in connecting the dots from a variety of sources.
Choosing Sensors/Tags and Connectivity
Depending on the application, different location technologies are appropriate.
Using RFID tags makes it possible to create a database of materials, equipment, tools, and other tagged assets.
BLE beacons offer an opportunity to develop proximity-based applications or build location accuracy profiles without having to equip every asset with GPS.
UWB opens up the possibility of using high-precision location data where it is needed.
GPS / GNSS is more suitable for outdoor use and is appropriate for moving objects in general.
The key is to choose technologies based on requirements, not on capabilities.
Accuracy, range, and operating environment are important factors, as are power supply and economic considerations.
Edge Processing
Not all processing should happen in the cloud.
Some analytics can take place at the edge.
By way of example, a gateway or edge server can collect information from local sensors and connected devices, filter it, and send notifications to the cloud.
Instead of sending all the data in the system to the cloud for processing, an edge server can analyze it, identify an anomalous condition, and notify the operator.
Such a system reduces the amount of data coming from the system and allows you to respond to problems faster.
Connect AIoT Data to Your Existing Systems
One of the challenges for construction companies adopting IoT is that they are using a variety of enterprise resource planning systems and business management software.
Connecting these systems together is why BIM platforms such as Procore, Autodesk Construction Cloud, and others are so attractive: because they allow you to consolidate data from different systems under one umbrella.
An isolated IoT system represents another data silo that does not interact with anything else.
A better approach would be to connect relevant information from IoT systems with other enterprise data.
For example, such an architecture could look like this:
IoT Data + Asset Data + Project Data + Other Data = Analytics Power
By way of example, an event with the location of a particular asset becomes much more interesting when combined with information about what this asset is, what project it is on, what other activities are being performed there, and how this equipment has been used in the past.
This helps to better understand the value of each particular event.
A similar approach can be used in analyzing patterns of equipment use, working hours, and comparing them with project requirements.
This is where the value of using AIoT in construction lies.
Where Does AI Come In?
By the way, AI is not something magical that must be used in every project.
Many IoT applications do not require AI processing; event-triggered notifications or simple rule systems are sufficient.
AI is useful when you have enough data for correlations to be found that were not apparent at the time of writing the code.
Practical applications could include:
• Anomaly detection
• Equipment utilization analysis
• Change detection
• Construction progress analysis
• Combined analysis of multiple data sets
• Finding correlations between location data and project events
This is an important qualification: AIoT is not always necessary for a particular application, and even if it is, it is useless without the required data infrastructure.
Data Quality Is Often More Important Than Algorithms
Even the best AIoT system will fail if the data is incorrectly formatted, incompletely transmitted, or completely unavailable.
A practical implementation should therefore take a realistic look at data quality.
Some questions to ask include:
Is there a sufficient number of uniquely identifiable assets?
How often do I need to update location data?
What degree of positioning accuracy is required?
What happens if a particular asset is out of radio range?
Where is it better to process data: at the edge or in the cloud?
Which events need to trigger an immediate reaction?
What information needs to be stored for subsequent analysis?
What systems should this data be integrated with?
By answering questions like this, you can design a complete architecture that will serve the needs of your business.
An Overview of the Proposed Architecture
A simplified version of the described architecture might look like this:
Construction Site
↓
Sensors / RFID / BLE / UWB / GPS
↓
Gateways & Edge Processing
↓
Connectivity / APIs
↓
Data Platform
↓
Analytics / AI
↓
Construction & Enterprise Systems
↓
Operational Decisions
Depending on the use case, this architecture may vary, but these are the main elements of the system.
For those interested in learning more about the specifics of implementing AIoT in commercial construction, CommCon AI is an interesting resource that is focused exclusively on construction.
Summary
AIoT in construction is not about random sensors; it is about connecting the dots between physical assets and digital data.
The biggest wins will come from clearly defined use cases that seek to answer specific questions and understand what data is needed to address them.
The choice of particular technologies (connectivity, edge processing) will be based on these goals and data realities.
This is why AIoT is a great idea for building a connected construction ecosystem that serves a clear business purpose.
Top comments (0)