טכניון מכון טכנולוגי לישראל
הטכניון מכון טכנולוגי לישראל - בית הספר ללימודי מוסמכים  
M.Sc Thesis
M.Sc StudentNaseraldin Hussein
SubjectIntegrated Planning of Product Design, Assembly Process and
Supply Chain
DepartmentDepartment of Industrial Engineering and Management
Supervisors Professor Boaz Golany
Professor Eyal Zussman


Abstract

Competition in the marketplace forces industrial organizations to concentrate on process improvement and time reduction.  A critical process to be improved is the decision-making process that links various design and planning activities.  Typically this process is done in a serial manner.  The process starts with designing the product bearing in mind only design aspects.  The selected design is then transferred to the planning function in order to determine the assembly sequence, subject to the selection of the design made earlier.  The selection is based merely on planning considerations.  Both decisions become constraints for the logistics function when designing the supply chain.  Solutions provided by the serial process are classified as local optimum, i.e., they are not necessarily effective in a global perspective.  In order to improve this decision-making process, we propose a simultaneous pattern.  Basically, this simultaneous pattern is an implementation of concurrent engineering fundamentals.

A great body of research has been conducted in the area of concurrent engineering.  Only a small portion of it incorporated quantitative aspects into the design, assembly planning and supply chain functions.  In this research, a quantitative model based on a Goal Programming technique is presented.  The model incorporates considerations drawn from design, assembly planning and supply chain.  The outcome is a compromise solution that satisfies the various aspects relatively to a preference data set.  The research simulation results show that the simultaneous process is far better than the serial process or any compromise solution that is randomly generated.

Sensitivity analysis of the model reveals that it is rather sensitive to aspiration levels and weights assigned to its goals.  In order to overcome this sensitivity, specific interactive algorithms are presented.