Hal,

"The future of project controls is to place the responsibility for gauging where we are and how we are doing at every moment with the people who are executing the project. No separation of execution and control. "
Is the antithesis of how aerospace and defense operate. From my experience there and in ERP and Enterprise software, this is the literal kiss of death of a project. Those with the technical responsibility have very little if any incentive of reporting accurate and credible progress to plan. The very nature of planning and controls is to have an independent and credible assessment of the physical percent complete.

I’d be VERY interested to see how this notion has come about in your construction domain? This is not to say that the technical staff does not plan, concur with the measures of progress – percent complete, apportioned milestones, physical deliverables, or the myriad of other assessment of progress. But to have those performing the work to be responsible for reporting the “gauging” the progress is strictly forbidden in government contracting – for all the right reasons. Those responsible for delivery have a disincentive to report bad news.

Phil,
From my experience (limited to ERP, defense and large constructure), the core issue is defining what “done” looks like before the work starts, measuring progress toward “done” in some unit meaningful to the participants (customer and supplier), and agreeing what needs to take place in the work efforts when “done” is not appearing at the planned cost, schedule, and techncial performance.
This is the primary rols of Program Controls- to provide this “actionable insight” to the project management and technical staff, so they can in fact keep moving toward “done” at the planned rate.
I’m having a hard time accepting that those doing the work are the right people to be telling management they are making progress. Over and over again, I’ve seen troubled projects where there was no seperation between “doing” and “reporting.” Everwhere from Joint Strike Fighter, to SBIRS (Space Based Infrared System), to possibly the Big Dig, to the IRS $4B rightoff, to the classic ERP deployment where “we’ll make it up in system test” becomes the watch word. There are dozens of GAO reports that point out this failure mode.
Kent Beck had a great quote. “Optimizim is the disease, feedback is the cure.” But this feedback must be a credible indepedent assessment of physical progress to plan. Otherwise your suggestion results – “we’re stuck at 80% complete for 150% of the time.”
I truely await Hal’s description of the details of how allowing those doing the work to be the only source of progress to plan.

The way I’d heard the old saw that Phil quotes is:

“a task is 90% complete for 90% of its duration….ask anyone”

Same principle: without objective measures, who can tell. For a lot of construction tasks tho we can measure: count bricks, measure volume of concrete placed, mass of steel erected, length of wall completed, number of windows installed; and deviations from promises are also dead easy to spot on most occasions: just count the items.

What makes Last Planner work is perhaps the next trade contractor expecting to take up their work by a promised date, with the previous one possibly under financial incentive to meet the commitment (liquidated damages for non-completion).

In the NSW Govt. standard GC21 contract, there is an early completion bonus that can be provided which also creates an incentive for the head contractor to meet its ‘defect free’ completion ahead of time.

I suspect that as often is the case with innovative thought, that babies go with bathwater. The benefits of what I’d call ‘humanistic’ project management: my term for the approach espoused by Hal and others, should be measurable. In my work in construction in the NSW Govt. we did find that by one measure our cooperative approach to contrating reaped huge rewards: contract litigation plummeted; but that was influenced by a comprehensive systematic change in our procurement approach.
But I’d expect that if humanistic approaches produced better results they’d be measurable generally; EV is probably a good method on very large projects with adequate controls; the challenge on smaller projects is to find a control method that allows simple communication to sponsors, provides useful information to enable corrective action to the project manager and team, and has predictive value, without denying that the highly communicative demands that humanistic methods entail.
I think the difference between humanistic and mechanistic PM approaches are much like those between modern cooperative general management and taylorist ’smoke stack’ management approaches in style; but both styles still need accurate book keeping.

Hal:

I think you are really on to something here. People closest to the work have the best insight, and the truth about what is and isn’t working, is or isn’t getting done, etc. I have used this approach, successfully, with a few additional extensions to this basic idea. One: It is difficult to establish partial progress (percent complete) but relatively easy to establish done, or not done. Two: For activities (tasks) in succession, the next step validates the previous (can’t hang drywall if the studs aren’t there) so, the concept of self reporting isn’t as irresponsible as it might first appear (reference to other posts here). Last, if you allow automated reporting of done or not done, within a dynamic updating environment, the methodology will scale to a project of any size, and, allow for more detailed task assignments than would be practical if progress updating is by manual means. So, make the plan very detailed, assign each piece uniquely, allow self reporting with dynamic calculation of earned value and (time) estimate to complete. With only 100 tasks reporting, you will get live earned value measurement to a resolution of approximately 1% of total (Tasks are never identical and of equal effort). That’s better than you will ever get estimating, and, there is no cost or time delay in getting the data. I’ve done this on knowledge work projects of all types, 200-500 team members wide, and it has worked flawlessly. I’d be interested in your views on this approach. The traditional Project Controls paradigm has been perpetuated within engineering and construction for some time, but I believe there is an opportunity to rethink the approach, as outlined above.