Cultural change for LSS, was that tough for your team?
Implementing any new process at an organization has some challenges. Incorporating the tenets of Lean Six Sigma (LSS) into the cost estimating and analysis process at the U.S. Army’s Program Executive Office for Simulation, Training and Instrumentation (PEO STRI) was made much easier since senior leaders supported the process and the expected outcomes from performing LSS projects. Given that the key goal of LSS is to improve organization processes resulting in cost avoidance or savings, it’s not hard to imagine that managers would get behind it. Certainly, deploying LSS at an organization takes time. This is because folks must be trained on how to complete a LSS project, and participants must learn how to help out the LSS team doing a project. Once the deployment is complete and some projects start to be completed, the organization starts seeing the benefits, and from there it’s only a matter of time until the organization’s culture is changed into a “lean” culture!
You started to talk about TrueFindings® . Can you explain the application further?
TrueFindings® is the data management application in the PRICE Cost Analytics suite that essentially is the companion statistical package to TruePlanning®. TrueFindings® enables analysts to perform data normalization, data analysis, and data management, with a focus on incorporating historical data analysis into an ongoing estimate to increase validity and ability to defend an estimate. Using TrueFindings®, analysts can create “findings” through data analysis; link findings directly to a TruePlanning® estimate; view descriptive statistics of data distribution; calculate correlation between variables; and develop bivariate and multivariate regression equations.
You mentioned that LSS DMAIC projects are “gated”.
Tollgate reviews are conducted by Project Sponsors and occur at the end of each phase of a gated DMAIC project. These tollgate reviews are intended to:
What were the major expected outcomes of implementing LSS in the Cost Analysis Division at PEO STRI?
There were three major outcomes that we wanted to see. The first and most important outcome was finding a solution to arbitrary and high growth factors applied to software cost estimates, the root cause of which was producing software cost estimates based solely on subject matter expert (SME) opinion rather than contractor software development engineering data (development code counts & development manhours). The second thing we wanted to do was to develop a standard process to gather SDE data, store it in a database, and analyze it properly to build good software development cost estimating relationships (CERs) across multiple domains. This database and the CERs could then be monitored, updated, and used for future estimating. Finally, we wanted to improve the quality and speed of conducting all software cost estimates. The PRICE Cost Analytics suite enabled us to do just that.
What were the major changes you had to make to reach your goals for the future of software estimating at PEO STRI?
There were several things we had to do to improve the process of software cost estimating at PEO STRI. The first thing we did was determine how the total process would work. This involved making it easy for a contractor to provide the SDE data (code counts & manhours) associated with a development effort. Then we developed a contract CDRL; comprehensive data dictionary; and user friendly code count and manhour capture methodology. In our case, we provided the developers with the free USC code counting tool and gave them a process to get the results sent to a standard form for input into our database. We also changed up our cost estimating processes to make use of the power of the PRICE Cost Analytics suite, where we could load and analyze SDE data in TrueFindings® and calibrate software cost objects in TruePlanning® using actuals rather than SME opinions. Finally, we had to update the cost estimating standard operating procedure to ensure all stakeholders complied with the new processes and that internal controls were maintained with respect to gathering contractor SDE data.