Shanthi Muthuswamy

Discrete particle swarm optimization for the team orienteering problem

Orienteering problem is a well researched routing problem which is a generalization of the traveling salesman problem. Team orienteering problem (TOP) is the extended version of the orienteering problem with more than one member in the team. In this paper the first known discrete particle swarm optimization (DPSO) algorithm has been developed for 2, 3 and 4-member TOP. In the DPSO meta-heuristic novel methods have been introduced for the initial particle generation process. Reduced variable neighborhood search and 2-opt were applied as the local search tools. The efficacy of the algorithm was tested using seven commonly used benchmark problem sets ranging in size from 21 to 102 nodes. The results of the DPSO algorithm were compared against seven other heuristic algorithms that have been developed for TOP. It was concluded that the developed DPSO algorithm for the TOP is competitive and robust across the benchmark problem sets.

Lean methodologies to improve assembly line efficiency: an industrial application

Similar to automobile industries, the large earth moving equipment (LEME) industries have a dominant assembly line layout. Improving the assembly line efficiency would improve the productivity of the line. This paper focuses on using lean tools such as time and motion study, kaizen events and spaghetti map to reduce waste in a LEME assembly line. For this research, five zones (or work areas) of assembly were studied and using the lean tools the cycle time was reduced to achieve a 15% reduction in takt time. There were four phases to this research effort: data collection, construction of the process flow map, creation of the spaghetti map, and the implementation of kaizen events. The results from this research demonstrate that waste in an assembly can be identified and removed by studying the process and utilising simple yet powerful lean tools.

Minimising makespan in a no-wait flow shop with two batch processing machines: a grasp algorithm

In a flow shop with two batch processing machines (BPMs) one of the key objectives is to minimise the makespan. In this study jobs with different sizes were batched together, without exceeding the machine capacity, and processed in the BPMs. Job ready times were also taken into consideration hence, a batch cannot be started unless all the jobs in the batch are ready and available. Also, the batches cannot wait between two machines (i.e., no-wait). The problem under study is NP-hard. A greedy randomised adaptive search procedure (GRASP) algorithm has been developed for this problem. Experiments were conducted to study up to 200 job instances. The solution quality of the GRASP algorithm was compared with the results of a mathematical model developed using CPLEX and a particle swarm optimisation algorithm. The experimental study highlights the advantages, in terms of solution quality of using GRASP to solve large-scale problems.

Scheduling batch processing machines in a no-wait flow shop using ACO to minimise the makespan

This paper considers a two-stage flow shop scheduling problem with a batch processing machine (BPM) in each stage. The processing time of the batch on the first machine is equal to the longest processing job in the batch, and the batch processing time on the second machine is equal to the sum of processing times of all the jobs in the batch. The jobs cannot wait between the two stages. The problem under study with the makespan objective is NP-hard. An ant colony optimisation (ACO) algorithm combined with batch forming and local search heuristics is proposed and its solution is compared with: a particle swarm optimisation (PSO) algorithm; a greedy randomised adaptive search procedure (GRASP) algorithm; and a commercial solver used to solve the mixed-integer linear formulation. The experimental study helps to highlight the advantages, in terms of solution quality and run time, of using ACO to solve large-scale problems.

A comparison of simulation and metamodelling approaches for production rate determination

Decision-making using simulation can be difficult, if day-to-day decisions have to be made using lengthy simulation runs. In situations where decisions are required quickly, an auxiliary model of a simulation model, called the metamodel, can be used. This metamodel is a regression model developed using the inputs and outputs of the simulation model developed for a problem. In this study, simulation and metamodels of a real-life production scenario were developed and validated for the purpose of comparing the two methodologies. The validation results justified that the metamodel could be used to estimate the response variable accurately and quickly. Along with modelling and analysing the present operation, four alternative scenarios were developed and validated to explore the possibility of improving the operation. The metamodels developed can be used for effective decision-making without the need for any specialised software or training.