A customer requirement identification approach was developed to achieve customer requirements and convert these requirements into engineering specifications. Quality function deployment (QFD) and fuzzy logic were used in this work.
A design modeling scheme was introduced to model variations of design configurations and design parameters. By defining the relations between designs (i.e., design configurations and design parameters) and design evaluations (i.e., functional performance and production costs), the optimal design configuration and its parameters can be achieved through genetic programming and constrained optimization.
A manufacturing planning, scheduling, and control method was introduced for one-of-a-kind production. Colored Petri-net was employed for simulation of this one-of-a-kind production system.
A method to predict manufacturing resource requirement based on
the prediction of the sales was introduced. Neural networks were
employed to model the non-linear relations between sales of different
products and requirements of manufacturing resources.
Three systems were developed based on the requirements from Gienow Windows and Doors, Calgary, Canada. These systems include (1) a Computer Aided Customer Interface (CACI) system to identify customer requirements and convert these requirements into engineering specifications, (2) a Computer Aided One-of-a-Kind Product Design (CAOKPD) system to model the generic families of OKP products and identify the customized products based on customer requirements, and (3) an Adaptive Production Scheduling (APS) system for planning, scheduling and control of the OKP production system. The manufacturing resource requirement prediction method was also used for predicting the labor requirements at Gienow Windows and Doors.
A manufacturing monitoring and control system was developed for Startec, Calgary, Canada. In this system, the design descriptions, modeled by components and assemblies of products, are associated with manufacturing descriptions, modeled by the operations in the production processes. Using the relations among design and manufacturing descriptions, the system can plan and report the status of operations, including “not-ready”, “ready”, “working-in-progress”, and “completed”.