Walt Scacchi
This research seeks to identify and describe a set of measures for assessing and improving the effectiveness of complex business or engineering processes at NAWCWPNS Procurement Section. This report seeks to investigate and demonstrate how concepts and computational mechanisms associated with process engineering technology developed at USC can be applied to help understand and measure what affects "procurement quality." The objective is to iteratively investigate, prototype and refine the initial set of process performance measures and computational methods that can be implemented and validated. An essential element of this research is the need to be able to have direct access to actual business or engineering processes within an organization such as MKPW for process modeling and performance analysis purposes.
This research seeks to iteratively identify and describe a set of measures for assessing and improving the effectiveness of procurement at NAWCWPNS. In this way, I seek to demonstrate how concepts associated with process engineering tools and techniques developed at USC can be applied in other business or engineering domains.
To help explain these concepts, I focus attention on the domain of procurement processes since every business or commercial organization has some form of procurement, acquisition, or purchasing processes and many have automated systems that support them. Similarly, procurement processes in some sense fall into the category of "overhead" for businesses whose principal activities or "core competencies" are in the manufacture, engineering, delivery or servicing of products, systems, or infrastructure.
In order to understand what affects procurement quality, there are three initial assumptions that underlie this measurement framework: (1) the population of procurement systems inputs and outputs (e.g., reports) that are consumed or produced within the procurement process are known or can be empirically determined; (2) that it is possible that NAWCWPNS can establish "Quality Circles" or the like for procurement staff to participate in; and (3) there is top management support and commitment of resources for the procurement process improvement initiatives. I believe all of these assumptions can be met at NAWCWPNS, as well as in other NAWCWPNS business units.
Next, there are three categories of measures. The first called "baseline measures" are of immediate concern. These measures seek to establish the basic level of process performance in a business organization. The remaining two, supplement the first to provide patterns of trends over time. These are called first-order and second-order measures, respectively. These measures are geared to collecting information on progress toward achieving process outcomes such as (a) customer satisfaction, (b) quality improvement, and (c) resource utilization. As such, both quantitative and qualitative data may need to be collected. Subsequently, we can identify a starting set of process performance measures, which we then seek to iteratively evolve through the analysis of observational and simulation-based process measurement data. These initial measures follow.
Measures in this category are of the same kinds as listed for first-order measures, but with one important revision. Second-order measures focus attention to the rate of change or "velocity" of the first-order measures, or, alternatively, the rate and direction of "acceleration" of the baseline measures. Of primary interest here are measures of (a) customer satisfaction, (b) quality improvement, and (c) total resource utilization to achieve (a) or (b).
Given the family of business or engineering process performance measures described above, we can now turn to highlight how these measures may be supported with computational mechanisms within an advanced process engineering environment.
At USC, we have developed and continue to use the Articulator process engineering environment over an eight year period. The Articulator is a knowledge-based environment that supports process engineering life cycle activities. These activities include process meta-modeling, modeling, analysis, simulation, tool integration, environment generation, monitoring and replay, articulation, and others. Using the process meta-modeling mechanisms, we can define a family of (static) process structure and (dynamic) process performance measurement objects, attributes, relations, and computational methods to augment the family of process models currently supported. Accordingly, we can then create instantiations of modeled processes using either empirical data on "as-is" processes, or hypothesize instantiation data for "to-be" processes. These instantiations can then be statically analyzed through computational methods that perform graph traversals, or dynamically analyzed through simulation or actual process execution. For simplicity, we will limit our initial research focus to the static and simulation-based dynamic analysis of processes which we can model. In turn, we can then iteratively model, analyze, and simulate both as-is and to-be business or engineering processes. Thus, the research to be performed here is to take the process performance measures, implement them within the Articulator, and then conduct simulation experiments and empirical observations to establish the basic validation of the process performance measurement framework.
Our research objective is to iteratively investigate, prototype and refine the initial set of process performance measures and computational methods to be implemented that we have identified above. An essential element of this research is the need to be able to have direct access to actual business or procurement processes within NAWCWPNS, for modeling and analysis purposes. As such, this project provides an ideal opportunity and supporting resources to conduct such a study, facilitated by NAWCWPNS, but for the mutual benefit of both NAWCWPNS and our research group at USC.