Construction sites are becoming increasingly connected.
RFID tags monitor material flow. GPS tracks equipment location. BLE devices offer proximity information. UWB enables accurate positioning. IoT sensors monitor environmental and equipment parameters. Project management systems store schedule and progress data.
The difficulty is no longer limited to capturing all these signals.
Engineering problem is to connect these diverse data sources together.
Once connected correctly, these signals can provide a highly valuable picture of what is going on on a construction site.
The Problem of Isolated Construction Data
Modern construction project can generate data from multiple sources.
For instance:
Workforce System
↓
Access/Location Data
Equipment System
↓
Telematics/GPS
Materials System
↓
RFID/Barcode/BLE
Project System
↓
Schedule/Progress
Each system might work flawlessly by itself.
However, most operational issues do not live within a single system.
Imagine a delay in execution of a particular activity.
Reason for delay could be:
Assigned work crew is not available
Necessary materials have not been delivered yet
Needed equipment is not available
Equipment is occupied somewhere else
Working area is not accessible
Previous activity was not performed
Reviewing just one data set does not give you a full picture of the situation.
This is where the problem of data integration comes into play.
AIoT: Connecting the Physical and Digital Worlds
AIoT is a combination of Artificial Intelligence and the Internet of Things.
Whereas IoT collects information from the physical world, AI helps to interpret it.
Architecture would be something like this:
Physical Jobsite
↓
Sensors / RFID / BLE / UWB / GPS
↓
Edge Devices & Gateways
↓
Data Integration Layer
↓
Operational Data Platform
↓
AI / Analytics
↓
Insights & Decisions
Actual architecture depends on specifics of a particular project, of course.
The task is to move from raw events to operational intelligence.
Why Project Context Is Important
Take an RFID event:
Material ID: M-2048
Event: Detected
Location: Zone B
Time: 10:42 AM
It tells you something about what happened at a certain time and place.
If you put it into the context of the project, it becomes very informative:
Material: Electrical equipment
Zone: B
Required for: Installation Task 42
Scheduled start: 11:00 AM
Crew: Electrical Team A
Equipment: Lift 7
In this case, the system potentially can check whether the required material is available before the planned activity starts.
Same goes for equipment and workforce.
Combining Multiple Events
One of the most interesting possibilities of construction AIoT is multi-source correlation.
Let's take such events as examples:
08:00 Crew enters Zone A
08:15 Equipment arrives
08:30 Material detected
08:45 Installation begins
10:30 Equipment becomes idle
10:35 Installation activity stops
A system based on equipment data alone could show something like this:
Equipment idle for 35 minutes.
But if we have all types of data on hand, the system could analyze a bigger picture.
Maybe the next required material has not arrived yet?
Maybe the crew is waiting for another trade to finish the preceding works?
There is a difference.
Good operational intelligence is not just an enumeration of events. It helps to understand relations between the events.
Where Does AI Fit Into the Picture
AI does not have to be responsible for all the decisions regarding construction works.
In a lot of cases, the best use of it would be to help understand patterns and prioritize attention.
Some potential applications of AI in construction include:
Anomaly detection
AI could help detect anomalies in equipment usage, movement patterns and operation as a whole.
Predictive Analytics
Historical data can be used to predict potential delays, equipment needs, or limitations in resources.
Pattern Recognition
AI can detect the relationship between the workforce, equipment, materials, and project data that might be hard to detect manually.
Decision Support
Instead of substituting for a project manager, AI can provide relevant information to allow a team to investigate the issue at hand.
Selecting the Correct Positioning Technology
There is no one solution when it comes to selecting the correct positioning technology for construction purposes.
GPS
Useful for outdoors, vehicles, fleets, and heavy machinery.
RFID
Used to identify materials, tools, components, and equipment when RFID-tagged devices pass certain predetermined detection points.
BLE
Used for proximity, visibility, and location-based services.
UWB
Provides precise positioning in environments where precision is key.
LoRaWAN
Depends on the specific deployment scenario; can be used for long-range low power IoT implementations.
Factors that should be taken into consideration when making your choice include:
Desired level of precision
Indoors or outdoors environment
Power supply considerations
Available infrastructure
Deployment cost
Frequency of data transmission
Specific physical conditions of the site
Thus, technology selection should start from the use case rather than the technology itself.
Edge Computing is Important As Well
Construction sites might not always have reliable connectivity, lots of data from sensors and latency sensitive applications.
Transmitting all events immediately to central cloud might not always be desirable.
Edge computing will make it possible to process information near its source.
Here is an example:
Sensor
↓
Edge Gateway
↓
Processing
↓
Relevant events
↓
Cloud System
A better design of construction intelligence platform pipeline might take into account whole data lifecycle:
Capture
↓
Normalize
↓
Integrate
↓
Store
↓
Analyze
↓
Contextualize
↓
Visualize
↓
Act
Normalization step is especially important.
Each system might have different representations of asset, worker, location, or event.
Here is an example:
System A: Equipment_007
System B: EXC-007
System C: Excavator-7
Without an identity model, linking these records together becomes a challenge.
An efficient data layer should provide unique identifiers and relations.
Security and Privacy Cannot Be Ignored
Connected construction spaces also raise security challenges.
The platform might receive data about workers, equipment locations, activities on a project, and the systems involved.
Some of the questions worth considering include:
Authentication
Authorization
Encryption
Data retention
Device security
API security
Network segmentation
Access auditing
Worker related data is especially sensitive.
Improving visibility doesn't necessarily have to result in constant surveillance.
Any deployment should have defined purposes, appropriate access controls, and transparency in place.
The Developer's Contribution to Smart Construction
Developing AIoT for construction is not as simple as calling an API with an AI model.
The developer has to think about:
Device connectivity
Event streaming
Data normalization
Entity resolution
Location systems
APIs
Edge processing
Cloud infrastructure
Analytics
Security
User interfaces
Most importantly, the technical architecture has to match the workflow on-site.
A brilliantly crafted data pipeline will not help much if the data arrives too late, it lacks context, or cannot be interpreted by the decision makers.
From Simple Data Collection to Operational Decisions
The biggest change when it comes to construction technologies is to move from simple tracking to actual decisions.
Knowing the location of an excavator can be useful.
Knowing the fact that it has been sitting idle for two hours is more valuable.
Knowing the reason behind this fact – that the required material has not been delivered yet – is even better.
And being able to know that earlier than the project manager does is when intelligent connections begin producing real value.
CommCon AI and other platforms are a part of the trend that is making connected AIoT, location, and construction possible.
The technology alone is one thing.
The value of the technology is in building connections between the physical world and operational context.
Final Words
The future of construction software is likely going to be connected environments that continuously exchange data among workers, equipment, materials, sensors, and project systems.
However, increased data flow does not guarantee a better construction project automatically.
The real problem for engineers is the creation of the system that turns the disparate physical-world signals into useful data.
AI can analyze the data. IoT can collect the data. Data engineering can link the data. And construction professionals can make the intelligence actionable.
This is the basis of a truly connected worksite.
Top comments (0)