Tapan P. Bagchi

Lot streaming and scheduling heuristics for m -machine no-wait flowshops

Abstract The objective of this paper is to minimize makespan in m -machine no-wait flowshops with multiple products requiring lot streaming. A ‘product’ here implies many identical items. ‘Lot streaming’ creates sublots to move the completed portion of a production lot to downstream machines so that machine operations can be overlapped. For the single product case with fixed number of sublots we obtain optimal continuous-sized sublots and then use a heuristic to find integer-sized sublots. For the multi-product continuous-sized sublots case we show that the optimal sequencing of products may be attained by solving a traveling salesman problem. We then construct another heuristic to yield integer-sized sublots. Finally, we evaluate the use of genetic algorithmic meta-heuristics for the interacting decision phases in simultaneous lot streaming and sequencing. We conclude that while GA may deliver makespans comparable in quality to those given by heuristic methods that cleverly exploit problem features particular to lot streaming, GA loses out in computational efficiency. On the other hand, GA can optimize the number of sublots for each product — a task for which neither an analytical nor a heuristic method presently exists.

A review of TSP based approaches for flowshop scheduling

Abstract Contemporary flowshops that are variants of the classical flowshop frequently pose challenging scheduling problems. Such flowshops include no-wait, blocking, and robotic flowshops. These may sometimes be modeled as traveling salesman problems (TSP) and then solved using efficient algorithms available for the TSP. Encountered in auto, electronic, chemical, steel and even modern service industries, such problems are reviewed in this paper. We show that the TSP based approach is quite effective over a broad range. It tackles no-wait flowshops, blocking flowshops, group scheduling of parts in a flowshop using a generalized extension of the TSP, lot streaming and scheduling problems, and as recently done, scheduling of parts and robot movements in automated production cells. In this review paper, we describe several well-documented applications of no-wait and blocking scheduling models and illustrate some ways in which the increasingly used modern manufacturing systems such as robotic cells may be modeled as TSP. We also review the computational complexity of a wide variety of flowshop models. Finally, we suggest some fruitful directions for future research.

Parto-Optimal Solutions for Multi-objective Production Scheduling Problems

This paper adapts metaheuristic methods to develop Pareto optimal solutions to multi-criteria production scheduling problems. Approach is inspired by enhanced versions of genetic algorithms. Method first extends the Nondominated Sorting Genetic Algorithm (NSGA), a method recently proposed to produce Pareto-optimal solutions to numerical multi-objective problems. Multi-criteria flowshop scheduling is addressed next. Multi-criteria job shop scheduling is subsequently examined. Lastly the multi-criteria open shop problem is solved. Final solutions to each are Pareto optimal. The paper concludes with a statistical comparison of the performance of the basic NSGA to NSGA augmented by elitist selection.

Numerical methods in Markov chains and Bulk queues

I: Introduction.- 1.1 A Perspective.- 1.2 Earlier Work.- II: Mathematical Formulation of the Bulk Queuing Problem.- 2.1 A Class of Markov Recurrence Relations.- 2.2 The Abstract Formalism.- III: A Numerical Approach to Waiting Line Problems.- 3.1 Numerical Methods in Queuing Theory.- 3.2 The Basic Theory.- 3.3 The Simple Queue with Limited Waiting Room.- 3.4 A Bulk Queue with Limited Waiting Room.- 3.5 Queues with Variable Arrival Rate.- 3.6 A Heuristic Experiment.- 3.7 Operational Solutions vs. Exact Results.- IV: Conclusions and Directions for Further Research.- References.- Appendix A: Program Listings.- Appendix B: On the Analysis of Computational Errors.- Appendix C: On Analytic Approximations.

Multi-Satellite Scheduling Using Genetic Algorithms

ISRO Telemetry Tracking and Command Network (ISTRAC) operate a fleet of Indian remote Sensing Satellites (IRS) in the low earth polar sun-synchronous orbits. Satellite Operations Scheduling of ground stations is one of the important tasks performed at Spacecraft Control Centre (SCC). In the prevailing multiple-satellite operations scenario, scheduling becomes complex, because of satellite-specific constraints, ground station configuration, satellite priorities and priorities of payload and special operations. Further, radio visibility conflicts are to be taken in to consideration, while generating the weekly operations schedule. Resolving the visibility clash at a ground station and allocating Telemetry, Tracking and Command (TTC) resources optimally for multiple satellites meeting the requirements of each mission are the key aspects of scheduling. Optimal resolution of visibility clashes is performed using Genetic Algorithms. The software, “Intelligent Multi-Objective Priority Activated Task Optimizer (IMPACT)” uses Artificial Intelligence and constraint directed search to generate weekly optimal schedule of spacecraft support. IMPACT is designed to resolve spacecraft controller clashes as well, thus enabling a single controller to perform multiple spacecraft operations. This software also ensures that the operations load is distributed uniformly to all the ground stations in the TTC network. This paper presents the multi-satellite operations scenario of Indian Remote Sensing satellites at ISTRAC and the details of operations scheduling scheme adopted. The variants of visibility clash topologies, the optimization scheme based on genetic algorithm including the mathematical model for gain function, scheduling software architecture and its salient features are described. Important design features of this operational software system are highlighted with examples.

Multiobjective Robust Design by Genetic Algorithms

Abstract One key feature of Japanese products is that their design quality is robust—it remains unaffected by the deleterious impact of environmental or other factors often beyond the users control. As a result, designing and producing such high-quality products have become the challenge for all manufacturers aspiring to be world-class in todays marketplace. This article shows how expressly crafted genetic algorithms can uncover and then exploit the dependency among the key design factors (the choice of materials, parts, configurations, manufacturing methods, etc.) and the final products performance, to obtain “robust” performance. The methodology presented extends single objective design prototyping procedures called Taguchi methods to multiobjective robust design problems—problems for which no convenient methods presently exist.

Near Optimal Ground Support in Multi-Spacecraft Missions: A GA Model and its Results

This study addresses the optimal allotment of ground station support time to low Earth orbit (LEO) spacecraft with clashing radio visibilities. LEOs now form a critical global infrastructure for natural resource management, rescue, crop yield estimation, flood control, communication, and space research and travel support. In the multi-spacecraft scenario, ground support becomes complex because of spacecraft-specific constraints, station configuration, spacecraft priorities and priorities of payload and special operations. A generalization of the classical product mix problem, spacecraft support is NP-complete and more complex than the former because of arbitrarily defined profitability profile. Genetic algorithms (GA) are used to near optimally resolve visibility clashes. It concludes with the illustration of real life spacecraft support optimization problems routinely faced by mission managers. A spin-off of this work is that it can enable the decision maker to also determine optimal ground station locations and support capability deployment in diverse planning scenarios.

Multiobjective Flowshop Scheduling

It was mentioned in Chapter 4 that in sequencing jobs in a flowshop we are likely to confront several different (and often conflicting) management objectives. Consequently, a schedule may have to be evaluated by different types of performance measures. Some of these measures may give importance to completion time (e.g. makespan), some to due date (e.g. mean tardiness, maximum tardiness), and some others to the speed with which the jobs flow (e.g. mean flow time). The simultaneous consideration of these objectives is a multiobjective optimization problem. But, even for a single objective, flowshop sequencing is NP-hard. Therefore, solving even a single objective flowshop problem involving only 15-20 machines and 30-40 jobs by classical optimization methods would be quite difficult.

The expenditure-GDP nexus: evidence from a panel of SAARC 7-countries

The paper investigates the casual relationship between government expenditure, GDP and exports for Bangladesh, Bhutan, India, Maldives, Nepal, Pakistan and Sri Lanka for the period between 1960 and 2010. Using cointegration and Granger causality, the results suggest that there exists bidirectional causality between exports and economic growth in India. For Bangladesh, Maldives, Nepal and Pakistan there is unidirectional causality from exports to economic growth. The unidirectional causality from government expenditure to economic growth is found in Bangladesh and Maldives, whereas the reverse causality is found in Bhutan and Pakistan. For Maldives and Sri Lanka Government expenditure causes exports, whereas for Pakistan export causes government expenditure. The panel analysis finally suggests and existence of unidirectional causality from export to government expenditure and from economic growth to government expenditure.

Growth, foreign investment and trade-openness interactions in ten OECD countries: a panel-VAR approach

This paper investigates statistical relationships between economic growth, foreign direct investment (FDI) and trade openness, using panel-VAR methods in relation to ten significant OECD countries: Austria, Canada, Finland, Iceland, Ireland, Japan, Norway, Spain, Switzerland and the USA, between 1970 and 2010. We find that foreign direct investment, trade openness and economic growth are co-integrated, suggesting the presence of long-run equilibrium relationships between them. Our Granger-causality tests indicate the presence of bidirectional causality linking foreign direct investment to trade openness, with unidirectional causality from foreign direct investment to economic growth, and also from trade openness to economic growth. Significantly, the pattern of causal relationships varies between the countries we studied.