Tuesday, January 26, 2016

CMMI for Service as Process Improvement

Capability maturity models answer the question: What are the characteristics of a high functioning organization? The defacto standard maturity model is managed by SEI ( and provides a detailed description of what highly mature organizations do.

As a process improvement tool, a CMMI model provides a standard against which the baseline organization can be compared. Any gaps between the standard and the baseline organization points the way for future improvement plans. The value of the model lies in the assessment material; it forces you to look across a broad array of processes and compare your organization to a consistent standard. Because the standard does not change, any reduction in gaps between the baseline and the standard represents progress for your organization.

The summary graphic shown here is my depiction of the CMMI for Service standard. It consists of 24 must-do processes that define a highly mature organization. Each abbreviation brick represents an important process area for a mature organization. As an example, the REQM brick represents requirements management.

Each brick in the pyramid is kind of like an individual assignment. You score the brick 0-3 based on how well your organization accomplishes that process. Then you combine scores on all of the bricks to determine your final grade -- your capability maturity level which is scored 1-5 (or F to A).

To conduct a formal CMMI assessment can be a monstrous undertaking that requires payment to outside assessors. In my organization, we are planning to start with an internal and informal assessment -- just to get us started and establish a baseline. We are the process now of translating the CMMI materials into focus questions for our organization.

Wednesday, December 9, 2015

DMADV for Travel Claims

The Navy's annual budget for Permanent Change of Station (PCS) moves is roughly $800 million. That sounds like a lot of money until you consider the scope and size of the effort. Between 110,000 and 160,000 PCS travel claims are processed each year. These claims include various allowances for time in training, family relocation, temporary lodging, and house hunting.

Because the Navy relies on rotational duty assignments by design and Sailors are entitled to PCS-related compensation by law, the expenses associated with PCS moves are a predictable cost of doing business.

However, the speed and accuracy of travel claim settlements has a significant impact on the operational availability of funds during the execution year. Adequate funds to safely cover all PSC-related expenses are obligated in advance of travel, and these funds must be held in abeyance until the travel claim is settled once travel is completed. Any excess obligations can then be de-obligated and used to fund additional PCS-moves. The goal is to settle travel claims within 30 days of travel completion. The process baseline in fiscal year 2015 was a median settlement time of 38 days. In addition, a small number of claims are never settled which ties up funds and creates the possibility of Sailor indebtedness to the Navy for advances paid on travel expenses.

We convened a meeting of the travel claim processing stakeholders in December of 2015 to discuss ways to (1) improve customer service, (2) improve the timeliness of travel claim settlements, and (3) improve the traceability of funds for the purposes of audit readiness. The group quickely settled on a DMADV methodology as an organizing framework for the effort. DMADV stands for Define, Measure, Analyze, Design, and Validate; it is a modified version of the familiar DMAIC model for lean six sigma process improvement. The DMADV framework is used when designing a new process, and DMAIC is more appropriate for refining an existing process.

After constructing a high-level conceptual diagram (see figure), the stakeholders designed a pilot test of the new process and constructed a Plan of Actions and Milestones (POAM) for execution of the pilot. The pilot test will run for a period of six months, includes a control group (i.e., all claims not in the pilot), and a thorough plan to evaluate the pilot against the measurable goals of the effort. If successful, the pilot will result in a Navy-wide implementation of the improved travel claim settlement process.

Tuesday, October 6, 2015

Telling a Good Story

I recently came across a fantastic news story in one of our process improvement efforts. Unfortunately, the effort and accomplishment was not fully appreciated -- due in part to the way the information was presented. Month after month, we presented the change as an unadorned bar chart. The bar chart showed clear improvements, but the sense of story was lacking.

After working with the team, we applied three recommendations to improve how the information was communicated.
1. We added control limits to the chart to convey a sense of context. How do we know when something changed? When the line breaks the control limits, something has changed.
2. We decided to show both the meter (the # of expired LIMDU PRDs) and the levers (the actions we had taken to influence change).
3. We made the math easy. Senior leadership no longer needs to calculate the size of the impact or guess; we spelled out the accomplishment explicitly.

Thursday, June 11, 2015

Measures of Performance: A Losing Record with a Winning Spirit

How do you measure performance in a recreational league?
Do you take into account bad umpires who don't seem to know the rules of the game?
Does your assessment take into account teams from competitive leagues who stay sharp in the off-season by playing in a recreational league?
Do the league mandated priorities of first fun, second learning, third winning play any role?

Although we have one game remaining, we will end this season with a losing record (currently, we are 1-7-1). However, we started the season with 5 of 14 players who had significant problems hitting a pitched ball. At our last game, every player on our roster got a hit. We started the season with 7 of 14 players who did not understand the basic flow of the game or how to get the other team out. At our last game, we held a very good team to two points in the first inning (a significant accomplishment for our team).

Everyone plays both infield and outfield. Everyone has improved. Our defensive play as a team has improved by leaps and bounds. And we're having fun. Although it is always more fun to win, we are enjoying the game of baseball.

Unfortunately, the scoreboard has been stubbornly resistant to measuring our true performance.

Friday, May 22, 2015

Building Demand for Black Belts

Because our deployment strategy has focused on developing bottom-up support for Continuous Process Improvement, we have been very deliberate about making gradual training investments. Initially, we focused on yellow belt and champion training with the role of "belt" for improvement projects supported from CPI program resources. As our trained population of yellow belts and champions grew, we began training green belts and broadening the base of belt leadership for projects. After three years of building, we felt it was finally time to invest in black belt training to expand the leadership team for the CPI program. This photo represents the first group to complete the 160 hour black belt training curriculum using our own instructors.

Tuesday, March 3, 2015

The Concept of "Process Entitlement" Drives Record Performance

During Lean Six Sigma Yellow Belt training, we introduce the concept of "process entitlement" as a way to drive record performance. Using a simple process simulation with building blocks to make pyramids, we lead a class discussion on what level of performance would be possible if the process was perfect. Perfect is defined as no waste, no variation, no constraints, etc. Collectively, we arrive at a general consensus of the natural physical limits of the process -- the maximum number of pyramids that could be constructed in a 5 minute production period. Then we steer the discussion toward the topic of, "What would have to be different to produce at a rate equivalent to the natural physical limit?" At some point during this discussion, we challenge the students with concepts such as Takt time and process metrics. Then, we reveal the current world record and set production goals somewhere above the current record and the maximum number possible. The record seems to inch higher with every new class and currently stands at 74 (with a competitor team in the same class producing 73).

Monday, December 8, 2014

Trained With An Expectation of Project Work

In November, we added 26 newly trained green belts to the program. All students entered the training with the expectation that they would follow through and complete two green belt projects to complete green belt certification. In theory, that should mean that 52 green belt projects will flow out of this one green belt class.

In reality, only about 40 percent of the students (10 of the 26, in this case) historically follow through to complete projects. Some of the reason for the project deficit are (1) the rotational nature of our workforce (i.e., green belts move on to new assignments before completing projects), (2) our "bottom up" deployment strategy that relies on green belts to generate project areas, and (3) our strategy of conducting projects with collateral duty green belts (i.e., project work is voluntary extra duty).