Customer order scheduling on two identical parallel machines with job setup times

Across all countries, manufacturers seek to adapt the best strategies to provide the highest qualities of services with lowest costs. For that matter, researchers have tried to develop better shop structures that in fact influenced and even can be controlled by a single machine, parallel machines and flow shop matters alongside with setup considerations. They especially put forwarded several approaches to scheduling by focusing on making certain alterations on setup timings of job assignments to achieve the best time saving and eventually the lowest cost. Studies on setup times or costs showed that, running the grouped jobs with the same or similar setup needs dramatically reduces both the setup times and costs. This procedure can also be classified as group technology and customer order scheduling (COS). The focus of this thesis is to provide an alternative scheduling for customer orders that may contain one or more than one job lots within two identical parallel machines. The common belief of any customer order is very straightforward which in fact requires its orders' to be proceed at the same time with a prompt attitude so that all job lot can be received with a well synchronization. The completion time of the last set of each customer order also indicates the completed duration of the customer order. Obviously, each job batch requires some certain setup arrangements unless setup needs of current job assignment match with upcoming assignment's setup needs. This study is an attempt to suggest a more time saving and low costing schedule by grouping and running customer orders with same setup requirements at most appropriate route via two identical parallel machines so as to reduce makespan time to present customers their orders in best optimized way. The existing problem has more than one variety; e.g., customer orders may contain more than one job assignment or orders may have different setup arrangements and times, etc. In such complex cases, it is fair to say that the problem is strongly NP-hard. MILP is practiced to solve optimal to small sized problems whereas a constructive algorithm is conducted to handle medium and large sized problems in order to get optimal and/or near-optimal solutions. GAMS, the optimization software for mathematical programming model is used to get optimum results. The heuristic algorithm is coded by computer language C++. In result of computational experiments, it was found that the mathematical model is inadequate to cover or may even fail to acquire solutions for especially medium and large sized problems. Keywords: Scheduling, Customer Order Scheduling, Group Scheduling, Makespan, Identical Parallel Machines