טכניון מכון טכנולוגי לישראל
הטכניון מכון טכנולוגי לישראל - בית הספר ללימודי מוסמכים  
Ph.D Thesis
Ph.D StudentGrossman Malki
SubjectApplying Analytic Tools of Probabilistic Risk Analysis as a
Method for Managing Development of High Technology
Products and for Improving their
Reliability
DepartmentDepartment of Industrial Engineering and Management
Supervisor Professor Emeritus Ishay Weissman
Full Thesis text - in Hebrew Full thesis text - Hebrew Version


Abstract

This research proposes controlled and quantitative Risk Analysis for Design Evaluation (RADE) for selecting design alternatives of high technology systems and products with relatively low risk. The tool is based on the structured methodology of Probabilistic Risk Analysis (PRA) and enables:

1.Product Knowledge

Each design alternative is organized into components for the creation of a database. Each component is characterized by type definition and its attributes (development complexity, component necessity and the required reliability during the development process).

The database enables the structured and gradual creation of a system dictionary, a common technical language.

2. Evaluation and Selection of Design Alternatives

Design alternative analysis is implemented by applying Fault Tree Analysis (FTA), and is based on the database created during the analysis’s first stage and includes the following steps:

a)      Analyzing Design Alternative:

1.   Modeling the system architecture with the SysML tool (if required).

2. Disassembling the system into components and characterizing each component.

3.   Building the Fault Tree of the design alternative.

4.  Semi-quantitative analysis of the system's fault tree: calculating the fault probability of the system design alternative (i.e. estimating the probability that the development process will fail). 

b)      Selecting the best design alternative (with the lower relative fault probability).

c)      Analyzing the selected alternative:

The Fault Tree can be applied to identify high risk components during the development process.

d)     Using RADE during the design process:

Sometimes there are several options for sub-system/assembly implementation. In such a case, the above mentioned a) and b) steps can be applied for deploying RADE as a decision-making tool.

The method has been tested through the analysis of several high-tech products, following the termination of their development process.

The analysis by RADE demonstrates that the highest estimated risk was, in actuality, represented by the selected alternative. In reality, the risk was realized, part of  the development processes were cancelled, and while others were completed, they entailed cost and time overruns.

RADE was tested and demonstrated only by "post mortem" analysis. All case studies had failed or been terminated unsuccessfully. It is recommended to apply the method on actual design processes.

Expanding RADE for other applications in the management of development process requires additional research. For example, work packages for system development can be defined, based on the database of system components, to improve schedule estimations of work packages, based on component attributes.