Mandyam M. Srinivasan

Tandem Configurations for Automated Guided Vehicle Systems and the Analysis of Single Vehicle Loops

Abstract Automated Guided Vehicle (AGV) systems continue to play a significant role in many low to medium flow manufacturing operations, including Flexible Manufacturing Systems (FMS) and other applications. The relatively inexpensive guidepath, coupled with the high degree of flexibility and control offered in vehicle routing, has made AGV systems a proven and viable handling technology for the 90s. Traditionally, AGV systems have been implemented and analyzed assuming that every vehicle is allowed to visit any pick up/deposit point in the system. We introduce a conceptually simple and intuitive approach where the system is decomposed into non-overlapping, single-vehicle loops operating in tandem. In this paper, we also develop an analytical model to study the throughput performance of a single vehicle loop. The resulting expressions are the first closed form analytical expressions that have been obtained to determine the throughput capacity of a single vehicle operating under a specific dispatching rul...

Tandem AGV systems: A partitioning algorithm and performance comparison with conventional AGV systems

In an earlier paper, Bozer and Srinivasan introduced the tandem concept for automated guided vehicle (AGV) systems and presented an analytical model to evaluate the throughput performance of a basic component of the system; namely, a single vehicle serving a set of workstations under the First-Encountered-First-Served rule. In this study, using the above analytical model and certain column generation techniques, we present a heuristic partitioning scheme to configure tandem AGV systems. The partitioning scheme is based on a variation of the well-known set partitioning problem. It is aimed at evenly distributing the workload among all the AGVs in the system. We demonstrate the procedure with two numerical examples. Using simulation, the performance of the tandem configuration obtained for each example is compared to that of the corresponding conventional AGV system.

Descendant set: an efficient approach for the analysis of polling systems

Polling systems have been used to model a large variety of applications and much research has been devoted to the derivation of efficient algorithms for computing the delay measures in these systems. Recent research efforts in this area, which have focused on the optimization of these systems, have raised the need for very efficient such algorithms. This work develops the descendant set approach as a general efficient algorithm for deriving all moments of customer delay (in particular, mean delay) in these systems. The method is applied to a very large variety of model variations, including: 1) The exhaustive and gated service policies, 2) Fractional service policies, 3) The cyclic visit order, 4) Arbitrary periodic visit orders (polling tables), and 5) Customer routing. For most of these variations the method significantly outperforms the algorithms commonly used today. >

TRIP-BASED MATERIAL HANDLING SYSTEMS: THROUGHPUT CAPACITY ANALYSIS

Abstract In this paper we present a general-purpose analytical model to compute die approximate throughput capacity of a trip-based material handling system used in a manufacturing setting. A wide variety of handling systems, including freight elevators, cranes, microload automated storage/retrieval (AS/R) systems, industrial lift trucks, and automated guided vehicle (AGV) systems can be modeled as trip-based handling systems. To our knowledge, this model is one of the few analytical models that explicitly considers an empty device dispatching rule. The model is first developed for a single-device system (such as a crane) and subsequently, with a simple modification, it is extended to multiple-device systems (such as lift trucks and AGVs). Using this model one can rapidly evaluate a wide range of handling and layout alternatives for given flow data. Hence, die model would be most effective when used early in the design phase to narrow down die set of alternative handling systems and configurations prior t...

Production-Inventory Systems with Preventive Maintenance

We consider a production facility that produces items for which demand occurs according to a Poisson process. The facility is assumed to deteriorate while it is in operation, with an increasing fai...

An approximation for mean waiting times in cyclic server systems with nonexhaustive service

The cyclic server system has been the subject of considerable research over the last few years. Interest in analyzing such systems has gained momentum due to their application in the performance analysis of token ring networks. In this paper we consider cyclic server systems with nonexhaustive service discipline. The performance measures of interest here are the mean waiting times at the nodes in the system. Exact analysis of such systems for these performance measures is very difficult in general, and a number of approximation schemes have been proposed in the past to evaluate these quantities. This paper presents a new approximation technique that gives accurate estimates of these mean waiting times, based on extensive validation with simulations.

Relating polling models with zero and nonzero switchover times

We consider a system ofN queues served by a single server in cyclic order. Each queue has its own distinct Poisson arrival stream and its own distinct general service-time distribution (asymmetric queues), and each queue has its own distinct distribution of switchover time (the time required for the server to travel from that queue to the next). We consider two versions of this classical polling model: In the first, which we refer to as the zero-switchover-times model, it is assumed that all switchover times are zero and the server stops traveling whenever the system becomes empty. In the second, which we refer to as the nonzero-switchover-times model, it is assumed that the sum of all switchover times in a cycle is nonzero and the server does not stop traveling when the system is empty. After providing a new analysis for the zero-switchover-times model, we obtain, for a host of service disciplines, transform results that completely characterize the relationship between the waiting times in these two, operationally-different, polling models. These results can be used to derive simple relations that express (all) waiting-time moments in the nonzero-switchover-times model in terms of those in the zero-switchover-times model. Our results, therefore, generalize corresponding results for the expected waiting times obtained recently by Fuhrmann [Queueing Systems 11 (1992) 109—120] and Cooper, Niu, and Srinivasan [to appear in Oper. Res.].

The relationship between preventive maintenance and manufacturing system performance

Abstract A common lament of the preventive maintenance (PM) crusaders is that production supervisors are often unwilling to lose valuable machine time when there are job waiting to be processed and do not assign high enough priority to PM. Maintenance activities that depend dynamically on system state are too complicated to implement and their overall impact on system performance, measured in terms of average tardiness or work-in-process (WIP) inventory, is difficult to predict. In this article, we present some easy to implement state-dependent PM policies that are consistent with the realities of production environment. We also develop polling models based analyses that could be used to obtain system performance metrics when such policies are implemented. We show that there are situations in which increased PM activity can lower total expected WIP (and overall tardiness) on its own, i.e., without accounting for the lower unplanned downtime. We also include examples that explain the interaction between duration of PM activity and switchover times. We identify cases in which a simple state-independent PM policy outperforms the more sophisticated state-dependent policies.

Expected waiting times in single-device trip-based material handling systems

Abstract In an earlier paper we presented an approximate analytical model to estimate the expected device utilization and the expected station cycle times (i.e., the average time between two successive arrivals of a device at each station) in a manufacturing system served by trip-based handling devices. Assuming that empty devices are dispatched according to the Modified First-Come-First-Served rule, the above model provides the expected device utilization, which the analyst can use to determine whether a proposed trip-based handling system is “stable”. In this paper we present an approximate analytical model to estimate the expected waiting times for move requests that occur in single-device trip-based handling systems such as cranes, vertical reciprocating conveyors, microload AS/RS, unit load Tandem AGVs, etc. The model represents a conceptual contribution, and it enhances the original model from a practitioners viewpoint since expected waiting times (and the associated mean queue lengths) can play an important role in deciding whether the performance of a “stable” system is actually “acceptable”.

Polling systems with state-dependent setup times

We analyze a polling system with multiple stations (queues) attended by a cycling server, in which a setup occurs only when the queue that is polled by the server has one or more customers present. Although such systems are appropriate for modeling numerous manufacturing and telecommunication systems, their analysis is not well developed in the literature. We provide an exact analysis for the 2 station model and present two approximation schemes to determine the mean station waiting times for models with 3 or more stations. We show that some approximate models which have been proposed in the literature for providing upper bounds on the mean station waiting times do not always yield upper bounds. Extensive numerical tests indicate that a simple average of the two approximation schemes yields a close estimate of the true mean station waiting time. This average-of-approximations procedure appears to be robust for a large range of parameter values.