Atomic Progress: The Smallest Details Lead to the Biggest Gains

TL;DR - New, disruptive technologies are being introduced to the construction industry that will leverage the power of visual-based modeling, artificial intelligence, and machine learning to streamline and improve the way we track project completion. Visual-based progress is anticipated to bring huge value as schedulers use cameras and other scanning technologies to automatically capture progress of a jobsite in a matter of seconds.

Looking for more? Sign up to my weekly newsletter here.

Listen to the only podcast dedicated to Construction Planning and Scheduling on Spotify , Apple Podcasts or YouTube .

A few years before I joined Google, I read an article about how Google trained a computer to recognize a cat . After I started working for the company, I had the opportunity to watch a demo of how they trained the computer to accomplish this feat. My mind was blown! Over and over, I watched the development teams leverage this capability and convert it into useful commercial products. Every time I saw the demos, I had a recurring thought in my head.

Construction must adopt these innovations.

Flash forward ten years, and the solutions have finally arrived in the form of visual-based progress. Today, there are a number of companies that offer some form of visual-based progress technology including AI Clearing , Avvir , Buildots , Constru , Disperse , Doxel , Openspace , and viAct (let me know if I missed any).?

“Disruptive technologies bring to a market a very different value proposition than had been available previously. Generally, disruptive technologies under perform established products in mainstream markets. But they have other features that a few fringe (and generally new) customers value. Products based on disruptive technologies are typically cheaper, simpler, smaller, and, frequently, more convenient to use.”

~ Clayton M. Christensen - The Innovator's Dilemma

How Does Visual-Based Progressing Work??

For those who have yet to experience a visual-based progressing service, the basic concept works like this:

• Step 1: Initial Job Walk - First, an individual or team is assigned to walk the jobsite with a special camera or LIDAR (Light Detecting and Ranging) device to capture the current progress of the project.?
• Step 2: Compare Captures to Existing Model - Next, the company uses the data gathered in the field to compare to a digital construction model to calculate the progress that has been made.?
• Step 3: Re-Capture the Jobsite - On a regular cadence or after a major milestone, the jobsite is re-walked to capture the latest progress.?

These services are quite amazing. They provide a decent ability to track progress against as-builts, clash detection, earned value, etc. Are they perfect? Of course not. But I don't want to talk about the here and now. I want to talk about where this technology could lead.

Where I Believe Visual-Based Progress is Headed

Here’s where I see the future going from my perspective. I imagine a world where almost any type of construction will be modeled quickly and everything that is needed to build a project will be produced quickly. To accomplish this, we’ll need technology that automatically creates detailed engineering, cost, scheduling, permits, pull plans, procurement logs, commissioning, and resource tracking.?

As this future job progresses, tools would seamlessly integrate with workflow processes (e.g., contracts), visual-based progress (e.g., steel erection), non-visual semi-automated process (e.g., cable pulls or system commissioning), near-location labor tracking (e.g., hours by electrical panel or time on tools), and precise equipment timing (e.g., fabrication, shipping, and installation. All of this while contractors and vendors are issued payments automatically based on the progress of the job.?

Not only would these scheduling features be interconnected, but they would leverage AI (artificial intelligence) and ML (machine learning) to assist human schedulers with optimization and predictions.

Do I have a crystal ball and know this for sure? No, but the unknown is rapidly becoming known.

Obstacles the Industry Needs to Overcome in a Visual-Based Progress World

A visual-based progress approach will be at the core of the future I envision. It will be the vehicle in which the model and schedule can be 1-to-1 at any level with limited operational overhead. In order to build an automated scheduling tool, I would focus on three components.?

• Allowable Physics - These tools need to be designed to be as realistic as possible. For example, you can’t build the second floor until you finish the first, you can’t fit three cranes within a certain area, you’ll need electricity prior to powering up equipment that requires electricity. If the program is designed with real life limitations in place, it will help avoid any issues with translating the virtual plan into real life.?
• Incorporating Best Practices - The tool would also need to know when and how to apply industry best practices such as starting east and moving west, finishing the floor before moving to fit-up instead of sections, or using two crews in an area instead of three.?
• The Engine That Builds It - I've seen quite a few demos of generative modeling. This would be an extension of that capability. A "Build The Schedule Button" that could be adjust using a variety of variables.

While these concepts are simplistic, they provide a solid framework to start from. The first item (allowable physics) is fairly straightforward and doable. Several of these capabilities already exist in one form or another. The second item (incorporating best practices) is also doable but would require an abundance of data on the minutiae of typical project work.?If you can nail the first two, putting a nice UI on the front would be the last step.

Since the beginning of time, we’ve had the ability to keep track of every nut and bolt of anything via analog means. The challenge is that these atomic components are an operational nightmare to track.

In an attempt to save time and effort, we often turn toward systems that aggregate the small details into more tangible chunks (think CPM, Lean, Takt, etc.). The problem is that the vast majority of these aggregations lose their direct connection to a specific piece in the model. They quickly become approximations which don’t help because we ultimately can’t identify what they are exactly.?

Enter visual based progress to save the day.

What Does This All Mean for Construction Schedulers??

Flashback to the present. What’s the point of using these services if they already don’t do all the magic you mentioned earlier? Using Clayton Christensen’s framework, I’ll let you decide… Does visual-based progress bring a different value proposition? Is it cheaper, simpler, or smaller? Is it more convenient to use?

I look forward to watching this technology grow and develop and sharing my thoughts with you.

Would you like to learn more about scheduling and get sent some of the best jobs around the world? Yes, sign me up!

Like the content but hate reading? Listen to our podcast on Spotify and Apple Podcasts and Youtube.

Does your resume need a fresh start? Check out my resume that has landed me top jobs here .

Need to prepare for an upcoming job interview? Here is a list of 26 of the best construction scheduler interview questions.