Project Controls and Work Sampling: A Potent Combination
October 24th, 2002 by GregEarlier notes on control were more conceptual. This one recounts how I became so suspicious of project controls. I became a consultant when I tried to sell Timelapse™ photographic equipment to contractors to analyze their construction operations. But they didn’t want equipment or to develop the ability to improve operations. They wanted someone to solve some specific problem. One of my first contracts was to help the project management team constructing a new coal fired power plant determine if the installation of the precipitator could be improved, and to figure out why data from two reporting systems was pointing in the opposite directions.
I arrived on site, introduced myself to the boilermakers, and climbed the chimney to record the operation. I watched it all day. The 6-man crew loaded 5 pipes into a skip and climbed the structure to where the skip was unloaded and the pipes installed. Then they climbed down and repeated the process. I knew after the first few cycles that they were spending less than 10% of their time installing pipes, and that another skip and small unloading platform would more than double performance. At 11:45, they left for a half hour lunch and came back at 12:45. This was all interesting and maybe useful. (Later I found the management really weren’t concerned about the precipitator installation process; rather they were sure the boiler makers were leaving early and getting back late.)
But the boilermakers weren’t the most interesting thing I saw. At first, I thought that prize would go to a group of laborers moving a stack of lumber, but by 2PM I was convinced it went to the work sampling program because I could always tell when a porta-potty was occupied.
Worksampling began 2 months earlier in response to declining productivity. Management was convinced workers were milking the job because they were lazy or dishonest. The rate of decline reflected in cost reports increased after the sampling began. Schedule slippage was growing too. Classic project controls showed things were getting worse, but reports from the sampling program showed workers were spending more time doing productive work. Workers were recorded as doing productive work if they were placing materials in their final position, and received half credit if they were moving materials. As a result, the laborers carried three of four pieces of lumber to the new stack and one on the way back. They had learned this improved the reports on their efforts. An occupied toilet was easy to spot because it had a short piece of pipe, lumber or rebar leaning next to the door. Everybody knew the reporting rules. Don’t go to the bathroom empty handed.
The exit interview was interesting. I showed the boilermakers were only doing "productive work" about 10% of the time, they took too long for lunch (one of them told me later, "You can’t get dinged for not working if you aren’t there."), and that a simple reorganization of work could double output. I wish I could say the project turned around because of this effort. They did drop the sampling program but they never could get past their belief that motivation was the problem. Not much happened to brag about. They never bought the photographic equipment – and even if they had, I doubt it would have made much difference.
Related Posts
- How do project controls work? We all know the way they are supposed to work. It goes like this -- A variance from standard is detected in the reports,...
- Preview of Linguistic Action Foundations of Lean Construction I've written about linguistic action as a theory that does a better job at explaining why projects succeed and why they ...
- The Future of Project Controls I'm somewhat hesitant to write about this. This morning I received an email newsletter that included advice on How to g...
- JM: Improving Work Systematically We've all heard about kaizen. It's a practice for individuals, teams, and across process. People seem to take to it ...
- Use a Gantt Chart as a Conversation Starter Project constituents always have two big questions: How is the project going? When will we finish? While the Ga...
