Keisuke Nagasawa

Robust flow shop scheduling with random processing times for reduction of peak power consumption

Abstract Recently, power shortages have become a major problem all over Japan, due to the Great East Japan Earthquake, which resulted in the shutdown of a nuclear power plant. As a consequence, production scheduling has become a problem for factories, due to considerations of the availability of electric power. For factories, the contract with the electric power company sets the maximum power demand for a unit period, and in order to minimize this, it is necessary to consider the peak power when scheduling production. There are conventional studies on flowshop scheduling with consideration of peak power. However, these studies did not consider fluctuations in the processing time. Because the actual processing time is not constant, there is an increase in the probability of simultaneous operations with multiple machines. If the probability of simultaneous operations is high, the probability of increasing the peak power is high. Thus, we consider inserting idle time (delay in inputting parts) into the schedule in order to reduce the likelihood of simultaneous operations. We consider a robust schedule that limits the peak power, in spite of an unexpected fluctuation in the processing time. However, when we insert idle time, the makespan gets longer, and the production efficiency decreases. Therefore, we performed simulations to investigate the optimal amount of idle time and the best point for inserting it. We propose a more robust production scheduling model that considers random processing times and the peak power consumption. The results of experiments show that the effectiveness of the schedule produced by the proposed method is superior to the initial schedule and to a schedule produced by another method. Thus, the use of random processing times can limit the peak power.

Semiconductor supply planning by considering transit options to take advantage of pre-productions and order cancellations

Abstract One of the objectives of supply planning is to find when and how many productions have to be started to minimize total cost. We aim to find the optimum. Base data like the length of transit time are important parameters to plan for the optimum start of production. In this research, we considered two kinds of transit options: normal transit and emergency transit, and we distinguished between planned and executed transit. The decision regarding which transit option to choose for the execution is trivial because emergency is only used when it is needed since normal transit is more cost efficient. However, for planning purpose, it is more difficult to decide which transit option should be used since the uncertainty in demand and supply between plan and execution can allow to plan an emergency transit but to execute the delivery with normal transit, which is a huge benefit in the competitive capital intensive semiconductor industry. If we planned an emergency, we can save inventory and production cost as we can delay the start of production. In contrast, we need pay additional transit cost in case that emergency transit is actually executed. Many characteristics of the semiconductor industry, which might be the front runner for many other industries, are considered in this model such as demand uncertainty, supply uncertainty and economic volatility. In our numerical experiments, we could gain the optimum, depending on each economic situation. Furthermore, we conducted sensitivity analysis of the effect of demand and supply uncertainties on total cost.

Consultation sequencing of a hospital with multiple service points using genetic programming

ABSTRACT A hospital with one consultation room operated by a physician and several examination rooms is investigated. Scheduled patients and walk-ins arrive at the hospital, each patient goes to the consultation room first, and some of them visit other service points before consulting the physician again. The objective function consists of the sum of three weighted average waiting times. The problem of sequencing patients for consultation is focused. To alleviate the stress of waiting, the consultation sequence is displayed. A dispatching rule is used to decide the sequence, and best rules are explored by genetic programming (GP). The simulation experiments indicate that the rules produced by GP can be reduced to simple permutations of queues, and the best permutation depends on the weight used in the objective function. This implies that a balanced allocation of waiting times can be achieved by ordering the priority among three queues.

DISASSEMBLY PLANNING BASED ON THE LINEAR INFLATION RULE UNDER STOCHASTIC DEMAND AND YIELD

The present study focuses on two stochastic elements, i.e., demand and yield, in planning the disassembly. The disassembly operation produces multiple types of parts, and the demand for each part type is generated independently. The present study applies the linear inflation rule (LIR) to the disassembly planning. The LIR defines the order quantity to minimize the expected cost by using a simple equation. The LIR assumes a single type of output. Thus the present study applies LIR to a selected part to obtain the quantity to be disassembled. This method may produce excess inventory or incur shortage in other parts. Thus it is allowed to dispose disassembled parts or purchase new parts adequately. As the selection of the maximum and minimum inventory levels greatly affects the performance, a single simulation run is applied, and the collected data are utilized to estimate the total cost under the different levels of maximum and minimum inventory. The effectiveness of the proposed method is confirmed through the numerical experiments.

CAN-ORDER POLICY WITH ITEM GROUP IN JOINT REPLENISHMENT PROBLEM

In this paper, multi-item inventory management is considered. The associated problem is called the joint replenishment problem (JRP). In this model, warehouse sells multiple items for customer and replenishes items from a supplier. For the replenishing, ordering cost is charged independent on the amount of orders. For this situation, can-order policy is often used. Under a can-order policy, items are ordered when their inventory level drops to or below their re-order level, and other item with an inventory level at or below its can-order level can be included in this order. In this situation, by considering can-order level and grouping the items can reduce ordering cost. In this research, setting of a can-order level and grouping are considered. The main objective in this model is minimizing the number of items lost sales cost, holding cost and ordering cost. In a numerical experiment, inventory movement are simulated.