Enabling Repeatable SE Cost Estimation with COSYSMO and MBSE

Abstract

In their paper on “A Generalized Systems Engineering Reuse Framework and Its Cost Estimating Relationships,” (Wang, Roedler, et al. 2014) present an approach for estimation of systems engineering effort that extends the COSYSMO equation to account for the effort associated with Design With Reuse and Design For Reuse classification categories in the Generalized Reuse Framework. Implementation of this approach for cost estimation clearly depends on two critical items: (1) the ability to accurately and consistently count the size drivers; and (2) the ability to calibrate the model equation. As part of future work, they also describe the potential to use this approach as a management tool in architecture development – “a cataloging mechanism in organizing components of reference architecture.” This paper presents a practical implementation of the COSYSMO cost estimating relationship through extension of a Model Based Systems Engineering (MBSE) modeling environment with SysML for estimating end‐to‐end systems engineering effort in developing a system. The approach provides a new way of rapidly creating cost estimates, conducting cost‐based analysis and trade studies with full traceability from the cost estimation parameters back to the architecture of referenced system of interest.

Integration of Parametric Cost Estimation with System Architecture – It's a dirty job but someone has to do it!

“…in which, k = {REQ, IF, ALG, SCN} Фx = quantity of "k" size driver, accounted for in the DWR process Ψx = quantity of "k" size driver, accounted for in the DFR process r = {New, Design Modified, Design Implemented, Adapted for Integration, Adopted for Integration, Managed} q = {No DFR, Conceptualized for Reuse, Designed for Reuse, Constructed for Reuse, Validated for Reuse} e = {Easy, Nominal, Difficult} wx = weight for "easy", "nominal", or "difficult" for the respective size driver wr = weight for defined DWR levels of the respective size driver wq = weight for defined DFR levels of the respective size driver (Papke, Wang, & Pavalkis, 2017), proposed a method of identifying and counting COSYSMO sizing parameters within the SysML modelling environment by taking advantage of the SysML language properties of profiles and stereotypes to extend the modeling language to include COSYSMO sizing elements. Given a system of interest (SOI), the architect identifies the relevant size drivers within the system architecture model and applies the corresponding COSYSMO sizing stereotypes as appropriate for their respective DWR and DFR categories and level of difficulty.…”

“…In our paper on implementation of parametric estimation with SysML and MBSE (Papke, Wang, & Pavalkis, 2017), we presented an approach that establishes the feasibility of integrating the COSYSMO parametric cost model with the SysML-based modelling environment so that key system architecture attributes can be automatically extracted as cost drivers and fed into the cost model to develop a cost estimate. This work is based on the initial description of the COSYSMO model (Valerdi, The Constructive Engineering Cost Model (COSYSMO), 2005) and subsequent work in 2014 by (Wang G. , Roedler, Pena, & Valerdi) and (Wang, The Generalized Reuse Framework -Strategies and the Decision Process for Planned Reuse, 2016) that extended COSYSMO with the General Reuse Framework to address cost estimating relationship for "design with reuse" and "design for reuse".…”