Ananth Krishnamurthy

Simulation based experimental design to identify factors affecting performance of AVS/RS

We perform a simulation based experimental design for automated unit-load (UL) storage and retrieval systems based on autonomous vehicle technology to identify factors affecting their performance. First, we select the best combination of numbers of lifts and vehicles from pre-defined scenarios that are the key components of the system. Then, we apply design of experiment (DOE) for a system with this combination of lifts and vehicles and for various arrival rates. The factors considered in the DOE include: dwell point policy, scheduling rule, input/output (I/O) locations and interleaving rule. Three different responses, the average cycle time for storage and retrieval transactions, average vehicle utilization, and average lift utilization, are considered. However, because the ANOVA assumptions are not met for the average cycle time response, an inverse transformation method is applied on this response. The results show that there is three-way interaction effect on each response at a 95% confidence level. After determining the main and the interaction effects, a Tukeys test analysis is completed on the responses. We utilize data from a warehouse in France that utilizes the autonomous vehicle storage and retrieval system.

Rechargeable sensor activation under temporally correlated events

Wireless sensor networks are often deployed to detect “interesting events” that are bound to show some degree of temporal correlation across their occurrences. Typically, sensors are heavily constrained in terms of energy, and thus energy usage at the sensors must be optimized for efficient operation of the sensor system. A key optimization question in such systems is—how the sensor (assumed to be rechargeable) should be activated in time so that the number of interesting events detected is maximized under the typical slow rate of recharge of the sensor. In this article, we consider the activation question for a single sensor, and pose it in a stochastic decision framework. The recharge-discharge dynamics of a rechargeable sensor node, along with temporal correlations in the event occurrences makes the optimal sensor activation question very challenging. Under complete state observability, we outline a deterministic, memoryless policy that is provably optimal. For the more practical scenario, where the inactive sensor may not have complete information about the state of event occurrences in the system, we comment on the structure of the deterministic, history-dependent optimal policy. We then develop a simple, deterministic, memoryless activation policy based upon energy balance and show that this policy achieves near-optimal performance under certain realistic assumptions. Finally, we show that an aggressive activation policy, in which the sensor activates itself at every possible opportunity, performs optimally only if events are uncorrelated.

Re-Examining the Performance of MRP and Kanban Material Control Strategies for Multi-Product Flexible Manufacturing Systems

Material control schemes can be classified as push, pull, or hybrid strategies. This article compares the performance of MRP (push) and kanban (pull) for a multi-stage, multi-product manufacturing system. Using simulation experiments we analyze system performance under different product mixes and observe that in certain environments with advance demand information kanban-based pull strategies can lead to significant inefficiencies. In these environments MRP-type push strategies yield better performance in terms of inventories and service levels. We also study the impact of design parameters such as safety lead time and safety stock policies on system performance and observe that for low and medium values of system loads, safety lead time policies yield better system performance than safety stock policies. These insights can be helpful in designing efficient MRP-type push strategies in multi-product environments.

Variance-based approximations of transaction waiting times in autonomous vehicle storage and retrieval systems

Approximation strategies for transaction waiting times are proposed for use in Autonomous Vehicle Storage and Retrieval Systems (AVS/RSs) with non-Poisson arrival rates and non-exponential service times. The proposed modelling tools dynamically adjust approximations based on the variance of the transaction inter-arrival times observed in a system. Simulation-based validation studies indicate that this strategy can reduce the magnitude of errors in estimated transaction waiting times by approximately 80% relative to static approximations. The design conceptualisation applications of the proposed technique are demonstrated over a series of realistically sized demonstration problems. [Received 8 April 2008; Revised 20 July 2008; Accepted 2 August 2008]

Planning and implementing POLCA: a card-based control system for high variety or custom engineered products

Many companies with high variety or custom engineered products are struggling to implement effective material control strategies on the shop floor, and finding that pull/Kanban systems are not meeting their needs in such environments. Paired-cell Overlapping Loops of Cards with Authorisation (POLCA) is a quick response manufacturing strategy designed with these situations in mind. POLCA is a hybrid push-pull strategy that combines the best features of card-based pull systems and push systems. At the same time POLCA gets around the limitations of pull systems in high variety or custom product environments. In partnership with its member companies, the Center for Quick Response Manufacturing (QRM) has recently implemented POLCA at several factories in the US and Canada. In this article, we first give an overview of the POLCA system, explain how it works and provide qualitative comparisons with pull/Kanban systems. Then, we present a step-wise procedure for implementing POLCA in a factory. Using examples from the implementation of POLCA at several factories, we address several practical issues such as computing the number of POLCA cards, determining the quantum of work a POLCA card represents, and addressing part shortages. We also discuss the different performance improvements that have resulted from these implementations including reductions in lead-time, increase in percentage of on-time deliveries and employee satisfaction.

Performance evaluation of a multi-product system under CONWIP control

Analytical models of multi-product manufacturing systems operating under CONWIP control are composed of closed queuing networks with synchronization stations. Under general assumptions, these queuing networks are hard to analyze exactly and therefore approximation methods must be used for performance evaluation. This research proposes a new approach based on parametric decomposition. Two-moment approximations are used to estimate the performance measures at individual stations. Subsequently, the traffic process parameters at the different stations are linked using stochastic transformation equations. The resulting set of non-linear equations is solved using an iterative algorithm to obtain estimates of key performance measures such as throughput, and mean queue lengths. Numerical studies indicate that the proposed method is computationally efficient and yields fairly accurate results when compared to simulation.

Performance analysis of single stage kanban controlled production systems using parametric decomposition

We present a new approach that permits efficient performance analysis of kanban systems with general demand processes, material arrival processes, and service times. The approach is based on parametric characterization of the traffic processes (arrival and departure) in the network and uses two-moment approximations to estimate performance measures at individual stations. We derive traffic flow constraints that are particular to closed queuing networks with synchronization stations and use these to establish relationships between the parameters characterizing arrival and departure processes at the stations in the network. The resultant set of non-linear equations is solved to estimate network performance measures. Numerical studies show that the approach is not only fast but also reasonably accurate when compared to simulation. These studies also provide insights with respect to the impact of different types of variability on the performance of a kanban system. This work also provides a fundamental building block that can be used in the analysis of multi-stage kanban systems.

Analytical models for analysis of automated warehouse material handling systems

Automated storage and retrieval systems (AS/RSs) and autonomous vehicle storage and retrieval systems (AVS/RSs) are two competing technologies for the handling, storage and retrieval of unit-loads in the reserve section of an automated warehouse. In this paper, we model variants of the two systems as an open queuing network (OQN) and use an existing tool for the analysis of OQNs, called the manufacturing system performance analyser (MPA), for analysing the performance of the AS/RS and AVS/RS. Experimental results are provided to show that MPA is a better choice than simulation to quickly evaluate alternate configurations of the two systems. We use MPA to answer a series of design questions for AS/RS and AVS/RS design conceptualisation.

Performance analysis and design trade-offs in warehouses with autonomous vehicle technology

Distribution centers have recently adopted Autonomous Vehicle-based Storage and Retrieval Systems (AVS/RSs) as a potential alternative to traditional automated storage and retrieval systems for processing unit-load operations. The autonomy of the vehicles in an AVS/RS provides a level of hardware sophistication, which can lead to the improvements in operation efficiency and flexibility that will be necessary in distribution centers of the future. However, in order to exploit the potential benefits of the technology, an AVS/RS must be designed using a detailed understanding of the underlying dynamics and performance trade-offs. Design decisions such as the configuration of aisles and columns, allocation of resources to zones, and vehicle assignment rules can have a significant impact on the performance of AVS/RSs. In this research, the performance impact of these design decisions is investigated using an analytical model. The system is modeled as a multi-class semi-open queuing network with class switching and a decomposition-based approach is developed to evaluate the system performance and obtain insights. Numerical studies provide various insights that could be useful in the design conceptualization of AVS/RSs.

Performance modelling of autonomous vehicle storage and retrieval systems using class-based storage policies

A cycle-time model for class-based storage policies is proposed for automated unit load Storage and Retrieval (S/R) systems that use autonomous vehicle technology. The model is based on a queuing network approach to achieve sufficient accuracy and computational efficiency for system design conceptualisation applications. It is illustrated through a series of realistic case problems.