Picture teams struggling to find the needed equipment, and managers attempting to rely on out of date information to manage a project.
Disconnected physical processes and digital information management may create conditions for these situations to occur.
Artificial Intelligence of Things (AIoT) addresses the problem by connecting Internet of Things (IoT) devices with AI to acquire and analyze information, which can improve operational decision making in the commercial construction industry.
Let's examine how this concept can be applied and what considerations should be made by developers and construction technology specialists when implementing it.
What Is AIoT in Construction?
IoT utilizes connected devices to collect information through sensors, tags, and other means.
AI analyzes this data to identify patterns and supports situational awareness.
In commercial construction applications, connected devices and technologies can be used to identify available equipment, track material locations and inventory, observe environmental conditions, or monitor machinery performance.
Depending on specific circumstances, a construction site may utilize a combination of the following:
RFID: Location tracking of tagged items
GPS: Long range tracking of tagged items
Bluetooth: Low Range tracking of tagged items
**Ultra-Wideband: **Precision tracking of tagged items
Environmental sensors: Temperature, humidity, and other conditions
Machine Telemetry: Operational information from assets
There is value in having an understanding of tagged asset locations, material inventory levels, and equipment operating conditions.
The specific technologies selected will depend on size and layout of an area, necessary precision of location information, available infrastructure, and budget.
The overall goal is to identify information helpful to making decisions and have an appropriate method to access it.
Improving Equipment and Material Visibility
Equipment and materials may require multiple transfers between storage, work site, and other areas of a construction job site.
If teams lack awareness of the location of assets, equipment operators may attempt to repurchase or rent an item that is already available at another worksite.
Material inventory issues can lead to ordering errors or confusion regarding the expected and received amounts.
Technology supported location tagging and tracking helps teams identify and address these situations.
Using a connected tracking system, a manager may verify if another worksite has the needed equipment before considering a rental option.
Similarly, tagged supplies may help clarify inventory problems.
Operators should be aware of the limitations of a tracking method before implementation.
For example, RFID requires a reader to be in proximity to a tag before it can be read.
GPS functions may be inconsistent in some areas.
Indoor positioning accuracy depends on specific infrastructure.
Predictive Maintenance Using AI
Construction equipment generates operational data that can be analyzed to support maintenance decisions.
Based on the type of equipment and sensors that have been installed, this information includes:
Equipment usage: Hours, patterns, etc.
Temperature: Operational and environmental
Vibration: Abnormal patterns, etc.
Status: Operational condition and error information
Using this data, AI can detect patterns that suggest a developing equipment issue.
As an example, an AI-powered maintenance support system might identify abnormal vibration patterns and alert personnel that an item may require inspection.
It is important to note that abnormal vibrations can be caused by a variety of factors, including poor maintenance, changes in environmental conditions, or sensor errors.
Before implementing predictive maintenance procedures, it is critical that AI has been trained on operational data and that maintenance personnel have been trained and are prepared to respond appropriately to alerts.
Job Site and Operational Safety with AI
A worksite involves the movement of equipment, materials, and personnel in a changing environment.
Connected sensors can help gather information on relevant conditions, and an AI system can be used to detect abnormal patterns.
As an example, an AI powered safety application may receive environmental data from a sensor, determine that the temperature is above a project's acceptable limit, and notify the appropriate team.
An appropriate response would involve identifying what action or inspection needs to be performed.
A properly functioning AIoT safety application would define these parameters:
What conditions will raise an alert
How to verify that an alert is valid
Who needs to be notified
What inspection or action should be taken
How to document the results of an inspection
With this information, a team can take appropriate action.
An AIoT application cannot assure job-site safety, but it can augment existing safety procedures.
Connecting Construction Data Systems
A complication of implementing AIoT is that location tags, inventory systems, equipment telemetry, and other data sources may exist in independent systems.
To unify this information, and bring it to an operator in a coherent manner, an AIoT system can be implemented using the following architecture:
- Devices
- Connectivity
- Edge Processing
- Data Platform
- Analytics
- Applications
Within this architecture, devices represent sensors, tags, equipment, or other sources of information that will be utilized by an application.
Connectivity refers to the system that allows devices to communicate with the remainder of the architecture.
Edge processing involves servers or applications that perform some calculations on device to reduce the amount of data that needs to be transferred, or to enable operations when a connection is unavailable.
A data platform is a database where data from devices and other sources is stored for use by applications.
Analytics software defines rules and algorithms to identify patterns in data.
Applications receive data from the data platform and present it to users, generate recommendations, or perform other functions.
Not all commercial construction projects will require all components.
A simple asset tracking application, for example, will likely involve a relatively straightforward architecture, compared to a project that will simultaneously track, analyze, and report on multiple data sources.
The choice of an appropriate architecture will be based on the needs of the application.
Sensors, Tags, and How to Use Their Information
It is one matter to collect information, and another to appropriately apply it.
As an example, consider an application that aims to improve equipment availability.
1. Collect: A tracking tag provides an update regarding the location and time of an item
2. Validate: The application confirms that it has received the update, validates that the update is recent, and may flag it as an issue if the tag indicates abnormal movement
3. Information: Based on this information, the application may calculate the likelihood that the needed item will be available
4. Recommendation: If the item is unavailable, the application can suggest an alternative based on current inventory and availability
5. Implementation: An operator reviews this recommendation and updates the inventory tracking information
As this example illustrates, an application utilizing AI should receive reliable information and generate reliable recommendations. Using AI as an intermediary between disparate systems and applications can help bring clarity and understanding to operators.
AI agents that assist with operational tasks need to be given responsibility that they can handle, and it is important that operators continue to review generated recommendations before taking action, especially in regards to physical equipment.
(https://commconai.com/)is an appropriate choice for organizations seeking an experienced team to assist with AIoT implementation in the commercial construction industry.
Challenges Developers Should Consider
Before pursuing an AIoT implementation initiative, it is important to understand potential obstacles.
Reliable Data Collection: Inconsistent, repeating, or irrelevant data may hinder an implementation.
Proper validation procedures should be established for each data source.
Connectivity: Large scale job sites may involve areas where a stable connection is not available.
Systems should be designed to buffer data until a connection can be re-established.
Integration: Older equipment or applications may not be compatible with modern systems.
Developers should identify integration points and work with available systems to determine compatibility and requirements before implementation.
Security: Connected devices and applications require secure protocols, including authorization and authentication procedures to prevent data breaches or sabotage.
*Alert Reliability: Excessive false alerts may diminish the importance of actual issues.
*
Thresholds and parameters should be established using appropriate rules, conditions, and AI models that have been tested using realistic data.
Information Accessibility: Data is only valuable if it can be interpreted and utilized.
Developers and operators should determine what information is needed and how it can be incorporated into existing procedures.
As a practical starting point, choose an obvious problem and verify that an identified solution actually helps.
A Practical Checklist Before Implementation
Before continuing an AIoT initiative, ask yourself the following questions:
What problem are you trying to solve? Consider a specific issue
What data do you need to solve it? Can your available sensors provide this information reliably?
What happens if a connection is lost? Can your operators continue to operate and verify the accuracy of alerts without this data?
What alerts and suggestions will you need to verify? Who has the authority and capability to perform these verifications?
What metrics will you use to evaluate the success of your implementation?
Final Thoughts
AIoT can help teams of commercial construction professionals identify and organize available equipment, track materials, and monitor environmental and machinery conditions.
By selecting appropriate sensors, tagging systems, and analyzing the resulting data, construction professionals can improve decision making and help avoid operational delays.
This process begins with understanding how to implement connected technology in a systematic, consistent manner.
For developers and construction technology professionals, the most appropriate point to begin is by selecting an issue or process that you would like to improve.
From here, you may design an appropriate architecture, incorporate existing systems or technologies, and generate data sources helpful to your operations.
Most importantly, people must be involved in information processing tasks that require context or discretion.
While technology can help monitor a job site, it cannot replace the value of individuals who make decisions based on knowledge, experience, and judgment.
Top comments (0)