سیستم های پشتیبانی تصمیم گیری برای عملیات تعمیر و نگهداری موثر

To compete successfully in the market place, leading manufacturing companies are pursuing effective maintenance operations. Existing computerized maintenance management systems (CMMS) can no longer meet the needs of dynamic maintenance operations. This paper describes newly developed decision support tools for effective maintenance operations: (1) data-driven short-term throughput bottleneck identification, (2) estimation of maintenance windows of opportunity, (3) prioritization of maintenance tasks, (4) joint production and maintenance scheduling systems, and (5) maintenance staff management. Mathematical algorithms and simulation tools are utilized to illustrate the concepts of these decision support systems. Results from real implementations in automotive manufacturing are presented to demonstrate the effectiveness of these tools.

In response to the challenges of fluctuating markets and the need for production of high volume of mixed products in a flexible manufacturing system (FMS), industry experts were surveyed for enabling technologies to improve the performance of flexible manufacturing. A survey was conducted by a CIRP Working Group on “Flexible Automation-Assessment and Future” in collaboration with the ERC for reconfigurable manufacturing systems during fall 2001 through summer 2002. The survey findings provide significant insights into the reasons for success and failure of FMS in the manufacturing industry. System capital cost was found to be the most critical factor in the success of large FMS. The cost of maintenance was the number two factor. The survey reveals that industry is dissatisfied with the high cost of maintenance of FMS and the actual system productive uptime is 25% lower than expected when installing the systems [1]. Generally, a large FMS contains many production machines, material handling and other pieces of equipment – all of which may break down during normal operation. These systems require (i) regularly scheduled maintenance or preventive maintenance (PM), (ii) repairs of machines reactive to the machine failures, and (iii) incidental maintenance tasks that require relatively small effort such as adding coolant or replacing tools. Appropriately coordinated maintenance-scheduling decisions can increase the system productivity if they are done based on information that is transferred comprehensively across hierarchical levels of control and management. Proper maintenance scheduling must consider both productivity and product quality. However, the mere use of maintenance scheduling is usually insufficient to obtain feasible solutions based on traditional combinatorial optimization methods because of the complexity of production processes [2]. Model-based maintenance decision support systems are needed to achieve high productivity and cost effectiveness of the overall system [3] and [4]. Decision making for effective maintenance of large systems is complex because it depends on several independent sources of information: (a) The Current health condition of each machine including: down; running; idling; being maintained; or just about to breakdown, (b) the scheduled daily, weekly and monthly maintenance plan, (c) machine health degradation profile, (d) throughput target and production rate, (e) costs of maintenance resource, i.e., labor, spare parts, tools, etc., and (f) the system configuration and decision alternatives [5]. Given the information from the machine and cell levels as inputs, the system-level controller is the best one for making effective maintenance decisions. These inputs then are compared to the overall production requirements that are sent down from the enterprise level. Design, control and management of such maintenance activities in large systems boost their productivities and increase their reliability and responsiveness to changing operations [1]. To this end, several important issues need to be addressed for the effective maintenance of large systems. They include (1) how to assess the impact of a machine breakdown on the factory throughput and determine what to do first, (2) if an unscheduled machine failure occurs, or if several events occur simultaneously, which reactive maintenance job has the highest priority (3) which machine failure is most seriously endangering the production schedule, (4) where are the opportunities for maintenance without affecting production throughput, and (5) how to efficiently utilize the factory resources (e.g., maintenance crews) on the critical sections of the systems. This paper addresses some of the questions raised above. We present several newly developed maintenance decision support tools for machining systems. We introduce throughput bottleneck detection and prediction techniques for guiding maintenance planning based on throughput-critical machines (Section 2) and a related method for estimating maintenance opportunities in a production line with machines and buffers (Section 3). From a joint production and maintenance perspective, we discuss maintenance tasks prioritization problem and an option-based model for maintenance scheduling (Sections 4 and 5, resp.). Section 6 presents a new decision support tool for effective maintenance personnel staffing management.

In this paper, we have described five important decision support tools for planning maintenance operations in an automation-based manufacturing system. These tools, aiming at providing intelligent joint production and maintenance decisions, will have a significant impact in three aspects: (1) efficient utilization of maintenance and production resources, (2) reduction in the unplanned downtime and hence improved productivity, and (3) minimization of the total cost of production operations. Some auto manufacturers have already benefited from these tools to improve operating efficiency and productivity, reduce total cost of production operation, and increase their competitive strength in rapid-changing markets.