Construction sites are physical places, but increasingly, they are also information sources.
People, equipment, and materials move through them, and project teams seek to capture this information for later analysis.
The challenge is to turn this flow of physical-domain events into something that software systems and humans can use — and that is where AIoT (the convergence of artificial intelligence and the IoT) can come in.
AIoT Defined
IoT (the Internet of Things) involves attaching some form of digital “intelligence” to physical objects.
In a construction context, this might involve using RFID, BLE, UWB, GPS, LoRaWAN, telematics, connected sensors, edge gateways, and other similar physical layer technologies to obtain information.
Meanwhile, AI involves building systems that can perform tasks that would otherwise require human intellect.
In AIoT construction applications, this means that rather than simply capturing data about equipment movements or other physical layer events, an analytics layer can analyze patterns within these data to identify trends or highlight items of interest.
The basic process involves:
Physical asset → sensor → connectivity → data pipeline → AIoT analytics → decision/action
Note that the value of AIoT lies in the data pipeline.
When discussing IoT, it is common to focus on the sensor.
While this is a critically important component, in most cases it is just one link in a longer chain, and one must think about the architecture as a whole.
A possible construction-sites AIoT architecture would involve:
Construction Site
↓
Sensors / Tags / Devices
↓
Connectivity
↓
Edge Gateway
↓
Data Ingestion
↓
Storage / Processing
↓
Analytics / AI
↓
Applications & Dashboards
↓
Operational Decisions
As can be seen, there are several layers that must be considered, and each has its engineering challenges.
The data must be reliable, the connectivity must be adequate, an edge gateway (when required) must be able to handle the data stream, the data ingestion process must work with the storage/processing layer, analytics/AI must be compatible with the dashboard/application layer, and the applications must be able to make recommendations or take actions based on the information received.
Real-World Construction Applications
There are three broad areas in which AIoT can be applied on construction sites:
1. Workforce Analytics
RFID, BLE, UWB, and GPS can be used to track individuals and provide information on their location, as well as other workforce-related metrics.
Depending on the tags and the associated readers/writers, this type of data can be leveraged to build workforce analytics.
The key challenge is that while the data is valuable, it must be aggregated and analyzed in context, taking into account privacy and other factors.
2. Equipment and Asset Tracking
Work equipment and assets can move around jobsites.
By using connected tags, it is possible to track their movements, and analyze patterns.
Once again, the value of the data comes from what the analytics can do with it.
For instance, equipment tracking data can be used to identify utilization trends, or abnormal movements.
This is more valuable than simply knowing the last known location of a particular asset.
3. Material Intelligence
Materials represent another source of information on construction sites.
By analyzing their movements, inventory status, delivery schedules, and more, it is possible to obtain valuable insights into their role in the built environment.
Linking these data to other analytics platforms can allow material intelligence to become much more valuable than it would be otherwise.
Core Engineering Considerations: Connected Data
Construction sites have many interconnected systems, including BIMs, ERP and CMMS systems, scheduling tools, procurement systems, project management systems, and more.
When designing an AIoT platform, it is critically important to remember that it must interoperate with these systems.
This requires that developers and system designers consider interoperability at every level, from APIs, data formats, and schema to event processing, device management, identity management, data quality, edge to cloud processing, cybersecurity, data retention, and more.
As the number of connected devices increases, so do the complexity and importance of these considerations.
Connected Construction Projects: Start With the Problem You Want to Solve
Many projects designed to leverage IoT data fail because they start from an unproductive viewpoint.
Developers and managers attempt to think about what they can do with a particular type of sensor, without first defining what it is they actually need to accomplish.
The correct approach is to define the problem first, and then determine what it will take to address it.
For instance, if a project team wants to be able to know where all of their equipment is, what it is being used for, and how it is being used, this represents a specific set of requirements.
From there, one can determine what combination of sensors, connectivity options, data pipeline components, processing/storage platforms, and analytics tools are needed to meet the requirements.
The point is to avoid getting caught up in trying to build an end-to-end IoT data platform without a specific purpose in mind.
From IoT to AI: Construction Insights and Analytics
The addition of AI transforms the potential value of IoT data, for construction analytics.
AIoT does not seek to replace human intellect, but rather to supplement it.
The most valuable opportunities are likely to involve linking the physical world (through IoT) to the digital realm (through AI).
IoT can provide answers to descriptive questions (“what happened?”), while AI can help look at patterns and determine what to do (“what does this mean?” and “what should we do next?”).
Ultimately, it is the latter that makes AIoT such a valuable field, for construction analytics.
The construction analytics industry needs more systems that seek to answer questions by linking the physical and digital worlds.
For developers seeking to build physical-digital systems, CommCon AI is a platform that brings together AI, IoT, tracking technologies, and construction to enable connected analytics.
Top comments (0)