Craig DeFinis enjoys watching a craftsperson lose 15 minutes of work time about as much as he likes discovering that he just left his wallet in the back seat of a taxi. As the owner of Pittsburgh plumbing and HVAC contractor DeFinis Mechanical Contractors, he grants his union plumbers a quarter-hour morning break, even though his contract doesn’t require it, and hopes the workers don’t stretch it the way the half-hour lunch break sometimes goes to 45 minutes. Because each worker costs about $60 an hour in wages and benefits, extra minutes add up over the long run. “Multiply that by eight guys, five days a week, for six months—that’s a lot of money,” he says.

DeFinis does more than gripe about breaks. He holds meetings with foremen every week to ensure timely delivery of tools, materials and information to his crews. Further, he stays alert to errors in the plans his crews work from. “Some engineers are better than others,” he says.

Overall, DeFinis reflects both the old and new ways of thinking about wasted work time. From the 1960s through the 1990s, owners, contractors and academics noted with frustration and alarm a steady lag in productivity within the construction industry, especially in periods of inflation. While technology had driven increased non-farm productivity ahead by 200% from 1964 to 2003, one controversial study showed, construction in the same period actually slipped by nearly 20%. Years ago, fingers may have pointed to union work rules and contract terms as factors, but a decline in trade-union market share in the last 30 years has changed that argument.

With a few exceptions, new research on productivity is giving managers the tools to maximize “wrench time.” Equipped with stopwatches and PCs to crunch their data, big owners, contractors and university researchers have turned to measuring and charting disruptions that can hit a jobsite daily. By drilling into the workday minute by minute, gauging output with new precision and zeroing in on ways to turn downtime into wrench time, researchers are cutting through the fog of confusion over what takes place on jobsites and, in the process, dispelling myths about lazy workers. Now, a major question is what to do with the rich harvest of activity data and how to exploit it consistently. A few measurement-minded companies are learning how.

In 2009, the Construction Industry Institute, based at the Cockrell School of Engineering at The University of Texas at Austin, launched the most ambitious study of craft productivity in its 28-year history. The five-year Craft Productivity Improvement Program aims to compile a comprehensive collection of innovative analysis techniques, common metrics and best practices for craft productivity.

The new research team, called RT 252, determined that project teams lacked a sufficient understanding of the issues that keep crafts workers off their tools.

Workers already had been polled on the subject. As part of a 2006 CII survey, 1,996 craftspersons listed what kept them waiting around: no forklift to move the Sheetrock, no reply from the engineer about unclear plans and specs, fixing prefabricated items and looking for tools, among other responses.

Productivity engineers argue that most contractors fail to investigate fully how these types of shortages and idle time disrupt production.

“Crafts guys want to be productive,” says Chris Buck, president of Productivity Enhancement Resources (PER) in Simpson, N.C. “You’re going to have your exception here and there, but the vast majority want to be productive. If we can take away the barriers for them to produce, they will take the reins and go.”

Buck is among a small group of productivity professionals engaged in the emerging technique of “activity analysis.” When it is hired by an employer or owner, Buck’s firm is sent to a site to make thousands of detailed observations about how crafts workers spend their time. While parsing the data and comparing it with historical norms, Buck looks for unusual spikes in non-direct work. He then goes back and interviews workers to find out why so much time is spent off-task. Once issues are identified, PER develops strategies to optimize the time of crafts workers.

The on-site observations produce mountains of data. Between March 2009 and February 2010, PER made 74,000 observations open-shop crafts labor on a Gulf Coast refinery project. The study led to an implementation plan that reduced total travel time for employees of an EPCM firm by 16% and cut operator idle time by 49% by taking downtime and using it for inspections of rigging equipment, cleaning of cranes and other machinery, and planning for lift coordination during the shift, says Beck.

PER then analyzed personal delays by the hour, showing a spike in the second half of the shift when temperatures were higher, he adds. So Beck advised supervisors to plan and schedule activities in the shade provided by existing structures, moving from one side of the project to another during the day.

The concept of observing direct and non-direct work rates isn’t new, but activity analysis represents an evolutionary step in productivity science. In the 1920s, Leonard Henry Caleb Tippett developed the “snap reading” method, which calculates the ratio between production and delay based on observations of work activity. Over time, some contractors began to gravitate toward the work-sample method, which splits a worker’s time into direct work, support work and delay.

Categories of Work

Through activity analysis, productivity engineers take sampling several steps further, breaking work into multiple layers that allow for more focused study. In July, CII’s RT 252 team released a comprehensive “Guide to Activity Analysis.” The guide suggests seven prime categories of work: direct work, preparatory work, materials handling, waiting, travel, personal, and tools and equipment (see above). Each category can be broken into subcategories to add detail.

To date, activity analysis has developed largely in silos in which companies have developed definition sets to meet their specific needs. Faithful + Gould, Seal Beach, Calif., established its own work-sample methodology, dubbed Time on Tools, in 2003. Dan Leng, vice president at the cost-management consultancy, says the system breaks non-direct work into nine main categories as well as smaller subcategories.

Although activity analysis appears laborious, Leng maintains that it is necessary in order to root out specific problems. Processing of permits, for example, can be broken down into many separate activities. “We break it down into type of permit, time waiting for someone to sign a permit, the signing process and so on. Eventually, you discover that a key person on the permitting process is constantly running 30 minutes late to sign a permit, for example.”

The payoff for micro-level information can be significant. Leng says his system saved $1.7 million by cutting logistics delays in half on a major turnaround job at a refinery, saved $835,000 by reducing breaks to 30 minutes from 55 minutes, and saved $330,000 by reducing by 10 minutes the owner “pep talk” during all-hands meetings.

“To me, it’s the most basic thing,” Leng adds. “If you have a project with 30% direct work, why would you spend time and money improving that direct work when you can try to fix the 70% [not spent on task]?”