Practical Application of Artificial Intelligence in Construction

A project manager wants a healthy project, yet few construction projects are delivered on time, on budget, with proper quality, and customer satisfaction. It sounds like a great way to apply Artificial Intelligence (AI), right? Track, monitor, and predict a Project’s Health automatically—in real-time—for thousands of projects simultaneously? Despite that AI is “eating the world”,  why is it not eating the construction industry as well?

What is a healthy project?

Construction projects can be defined as a sequence of unique activities, having one goal or purpose that must be completed by a specific time, within budget, and according to specification. Project performance is judged mainly on four performance metrics: cost, schedule, quality, and satisfaction. (Of course, Safety is always first!)

A good project manager will use experience and historical data to identify obstacles and warning signs in an ongoing project. They will take corrective actions or even proactively mitigate risks when starting or planning the project.

Fonn surveyed hundreds of project managers (who use Fonn software) to define the “healthiness” of thousands of completed projects. Fonn also collected millions of datapoints from tens of thousands of projects to reference and analyze, for the purpose of predicting project outcomes.

Using Artificial Intelligence, data can be compared between the completed projects, the survey, and the ongoing projects to provide real-time predictions on the health of active projects.

All projects are different

Construction projects differ in size, complexity, contractors, subs, clients, locations and many other tangible and intangible aspects. Based on user input and characteristics of each project, Fonn found a unique way of grouping and vectorizing projects to be able to compare current and completed projects for predictive analysis.

Representing project data

Each project in Fonn is represented with a dot or point in a graph. The closer each of the points are, the more similar the projects are. However, how can very large and complex construction projects be represented as a single dot on a graph? (Figure 1)

Project Vectors & Groups

There are a few steps. First, domain expertise and artificial intelligence is needed to identify distinctive features of construction projects. Next, algorithms are applied to quantitatively represent them for use in machine learning. Finally, we interpret what similarity and dissimilarity means for each feature separately as well as for the whole project. This process in AI or machine learning is called “representation, learning and clustering”. (Figure 2) We then use this data to compare characteristics of similar projects.

Representation, Learning & Clustering

Monitoring the “Heartrate” of a project

By enabling the use of data to evaluate and compare similar or different characteristics of projects, Fonn uses “Project Heartrate” as an intuitive tool for understanding a project’s “Health”.

So, how does a Project Heartrate work? It visualizes all the events happening within a project such as updates, deletes, discussions, schedules, tasks, and issues; and tracks them across the project. As we see in Figure 3, these communications are occurring with regularity within the project and are actively being documented. The results are an active, bright red “heartrate” indicator and subsequently a “healthy” project. That is, workers are actively addressing issues, collaborating on solutions, and remaining productive.

Heartrate of a

Alternatively, if project activity is reflected with a pale, anemically pink heartrate indicator (Figure 4), it can be identified as “unhealthy” (lack of communication, not documenting issues, incomplete tasks, etc.). It then serves as a guide for project managers to take measures to “defibrillate” and get the project back on track.

These “missing heartbeats” in the project are reflected in the documentation activity, checklists, issues, tasks, and lack of discussions. Perhaps subcontractors are not actively reading new revisions of drawings, or maybe documents are not even being revised and updated throughout the duration of the project. The Algorithm will highlight these deviations to let the project manager know which measures need to be taken to get the project back on track.

Steps to improve your projects

We have identified small steps your organization can easily implement to improve the “heartrate” and “health” of your projects:

1) Keep your project well documented. This is vital for a healthy project. Upload, revise, and share documents with all participants in your project to increase probability of it being successful.

2) Schedule your project—either tailored to your project or use an existing template. Report percentage completion and document completeness of your project to increase Quality.

3) Keep your conversations open and online. No emails, phone calls or simple messaging systems. Enrich discussions with the use of images. These are two small steps that increase Project Satisfaction and Quality.

4) Be pro-active! Projects that deactivate push-notifications are notoriously unhealthier than the ones that use notifications to actively keep the team up to date and informed.

5) Personnel Changes. Sometimes unavoidable, but changes of personnel and teams in projects seem equal in reducing Satisfaction.

What's next?

In short—start actively using a cloud based, user friendly project management tool today to have healthier projects in the future. Project management features (issues, tasks, schedules, document management) should be used throughout the duration of the project. Ensuring team members and contractors are not ignoring notifications and updates, as well as tracking crews through a site check-in feature is proven to have a positive impact on a project’s Health. Also, upload, revise, and share documents and drawings within the system. It enables you to propagate all changes and ensure that all members know the latest versions of documents.

Imagine being able to track your project progress and see indicators of potential problems as the project progresses! Imagine your construction AI assistant who helps you monitor your project’s Heartrate and Health! Let the Fonn team help you get started today!


image 16

Deliver your best work with Fonn

No credit card needed | Join over 500 others