A computer application for a bus body assembly line using Genetic Algorithms

This paper presents the implementation of a computer system in a bus body assembly line in Colombia, South America. The sequencing of the assembly line was formulated as a Resource Constraint Multi-Project Scheduling Problem. The proposed solution method was a classical Genetic Algorithm which was adapted to handle the constraints of the production system. The algorithm is embedded into a computer application that calculates the weekly schedule. The implementation of the computer system, in use since 2007, has brought dramatic changes in the company. The percentage of on-time deliveries rose from 65% to 85% in about one and a half years. The improvements have happened not only in terms of numbers, but also in the organizational climate at the company. The senior management frequently adds new functionalities to the application and the system is becoming a small Materials Requirements Planning (MRP). ► Application of Genetic Algorithms in real life with proven improvements over the current practice. ► Modeling of an assembly line as a multi-project resource constraint scheduling problem. ► Integration of a scheduling algorithm in a Decision Support System to manage an assembly line.