Systems are increasing in size and becoming more complex, an ongoing trend for decades. Systems monitoring and control is a vital element of systems engineering that provides information on the behavior patterns of the system. This information has various purposes, including, but not limited to, managing and improving the system.

Systems go through various lifecycle phases from concept definition to operations and maintenance (O&M). As O&M activities can cost as much as 75% of a product’s lifecycle cost, it is therefore important to effectively manage cost, optimize system “on” time, and mitigate defects/failures.

Statistical process control (SPC) in general, and control charts specifically, are the most widely used monitoring methods. The control chart provides alerts with respect to the behavior of systems and processes and changes in process variability, but relies on the normality of the underlying data. This constraint is easily satisfied in manufacturing and similar industries, where the natural variation in the process or system follows the Gaussian distribution.

Many systems today, particularly ones involving people processes and business rhythms, compromise the normality assumption and therefore appear to be poor candidates for SPC. This doctoral research focuses on the applicability and use of SPC in the O&M phase of a system where the analyzed data deviates from normality, for the express purpose of reducing the high costs by mitigating problems and uncovering inefficiencies. The target system is the Lockheed Martin Service Desk, and the challenges are the determination of the state of the process (in or out of control), and, additionally, when the process is out of control, which variable is the causing agent. Preliminary discoveries and results will be presented as part of this discussion.