Premium
Estimating Project Performance through a System Dynamics Learning Model
Author(s) -
Walworth Thomas,
Yearworth Mike,
Shrieves Laura,
Sillitto Hillary
Publication year - 2016
Publication title -
systems engineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.474
H-Index - 50
eISSN - 1520-6858
pISSN - 1098-1241
DOI - 10.1002/sys.21349
Subject(s) - rework , schedule , computer science , process (computing) , system dynamics , set (abstract data type) , project management , learning curve , process management , industrial engineering , systems engineering , operations research , engineering , artificial intelligence , embedded system , programming language , operating system
ABSTRACT Monitoring of the technical progression of projects is highly difficult, especially for complex projects where the current state may be obscured by the use of traditional project metrics. Late detection of technical problems leads to high resolution costs and delayed delivery of projects. To counter this, we report on the development of a updated technical metrics process designed to help ensure the on‐time delivery, to both cost and schedule, of high quality products by a U.K. Systems Engineering Company. Published best practice suggests the necessity of using planned parameter profiles crafted to support technical metrics; but these have proven difficult to create due to the variance in project types and noise within individual project systems. This paper presents research findings relevant to the creation of a model to help set valid planned parameter profiles for a diverse range of system engineering products; and in establishing how to help project users get meaningful use out of these planned parameter profiles. We present a solution using a System Dynamics (SD) model capable of generating suitable planned parameter profiles. The final validated and verified model overlays the idea of a learning “S‐curve” abstraction onto a rework cycle system archetype. Once applied in SD this matched the mental models of experienced engineering managers within the company, and triangulates with validated empirical data from within the literature. This has delivered three key benefits in practice: the development of a heuristic for understanding the work flow within projects, as a result of the interaction between a project learning system and defect discovery; the ability to produce morphologically accurate performance baselines for metrics; and an approach for enabling teams to generate benefit from the model via the use of problem structuring methodology.