Search results

The purpose of this paper is to investigate the performance of unpaced reliable production lines that are unbalanced in terms of their mean operation times, coefficients of variation and buffer capacities.

Simulations were carried out for five‐ and eight‐station lines with various buffer capacities and degrees of means imbalance. Throughput, idle time and average buffer level performance indicators were generated and statistically analysed.

The results show that an inverted bowl allocation of mean service times, combined with a bowl configuration for coefficients of variation and a decreasing order of buffer sizes results in higher throughput and lower idle times than a balanced line counterpart. In addition, considerable reductions in average inventory levels were consistently obtained when utilizing a configuration of progressively faster stations, coupled with a bowl‐shaped pattern for coefficients of variation and an ascending buffer size order.

The results for these specific experiments imply that resources expended on trying to achieve a balanced line could be better used by seizing upon possible enhanced performance via controlled mean time, variability and buffer imbalance. Results are valid for only the line type and parameter values used (simulation results are specific and not general).

Guidelines are provided on design strategies for allocating labour and capital unevenly in unpaced lines for better performance in terms of increased throughput or lowered idle time or average buffer levels.

This paper might be viewed as one of the first simulation investigations into the performance of unpaced production lines with three sources of imbalance.

The purpose of this paper is to investigate the performance of unpaced unreliable production lines that are deliberately unbalanced in terms of their coefficients of variation (CVs).

A series of simulation experiments were carried out for five and eight station lines with mean buffer space set at one, two, four and six units. CVs were allocated in 12 different configurations for each of these lines.

The results show that the best unbalanced CV patterns in terms of throughput rates or idle times as compared to a balanced line counterpart are those where the steadiest stations are concentrated near the centre of the line. On the other hand, either concentrating the steadier operators towards the centre or close to the end of the line gives best average buffer level results.

The results provide guidelines for production line managers when designing unpaced unbalanced lines depending on their performance aims.

The investigation of the effects of unbalancing CVs in unreliable lines has not previously been studied and can provide insights into how best to place workstations with differing variability along the line.

The purpose of this paper is to critically evaluate the state of the art of applications of organisational systematics and manufacturing cladistics in terms of strengths and weaknesses and introduce new generic cladistic and hierarchical classifications of discrete manufacturing systems. These classifications are the basis for a practical web-based expert system and diagnostic benchmarking tool.

There were two stages for the research methods, with eight re-iterative steps: one for theory building, using secondary and observational data, producing conceptual classifications; the second stage for theory testing and theory development, using quantitative data from 153 companies and 510 manufacturing systems, producing the final factual cladogram. Evolutionary relationships between 53 candidate manufacturing systems, using 13 characters with 84 states, are hypothesised and presented diagrammatically. The manufacturing systems are also organised in a hierarchical classification with 13 genera, 6 families and 3 orders under one class of discrete manufacturing.

This work addressed several weaknesses of current manufacturing cladistic classifications which include the lack of an explicit out-group comparison, limited conceptual cladogram development, limited use of characters and that previous classifications are specific to sectors. In order to correct these limitations, the paper first expands on previous work by producing a more generic manufacturing system classification. Second, it describes a novel web-based expert system for the practical application of the discrete manufacturing system.

The classifications form the basis for a practical web-based expert system and diagnostic benchmarking tool, but also have a novel use in an educational context as it simplifies and relationally organises extant manufacturing system knowledge.

The research employed a novel re-iterative methodology for both theory building, using observational data, producing the conceptual classification, and through theory testing developing the final factual cladogram that forms the basis for the practical web-based expert system and diagnostic tool.

The purpose of this paper is to develop a computer simulation model to evaluate the bowl phenomenon and the allocation at the end of the line of stations with either greater mean operation times or higher variability of operation times.

The model was developed on the basis of a realistic case problem and applied to a six‐station assembly line. The evaluation criteria were the: minimization of the total elapsed time; maximization of the average percentage of working time; and minimization of the average time in the system.

The performance of an assembly line with independently normally distributed operation times could be improved by applying the bowl phenomenon. The allocation of large operation mean times to stations located near the end of the line did not produce improved results. Instead a more balanced allocation proved to be more significantly effective. On the other hand, the assignment of larger variability of operation times to the stations near the end of the line improved the performance of the assembly line.

The investigation contributed to the computer simulation approach to solving assembly line problems that dealt with the impact of normally distributed operation times on the bowl phenomenon and assembly lines with increasing mean operation times and higher variability of operation times at the end of the line of stations.

The purpose of this paper is to introduce sequence‐dependent disassembly line balancing problem (SDDLBP) to the literature and propose an efficient metaheuristic solution methodology to this NP‐complete problem.

This manuscript utilizes a well‐proven metaheuristics solution methodology, namely, ant colony optimization, to address the problem.

Since SDDLBP is NP‐complete, finding an optimal balance becomes computationally prohibitive due to exponential growth of the solution space with the increase in the number of parts. The proposed methodology is very fast, generates (near) optimal solutions, preserves precedence requirements and is easy to implement.

Since development of cost effective and profitable disassembly systems is an important issue in end‐of‐life product treatment, every step towards improving disassembly line balancing brings us closer to cost savings and compelling practicality.

This paper introduces a new problem (SDDLBP) and an efficient solution to the literature.

Presents a model that minimizes an aggregate measure of waiting times and queue lengths for a collection of workstations. Shows that the naïve allocation of arrivals among workstations in proportion to their processing times does not result in an optimized system. Uses the Lagrange multiplier in determining the proportional distribution rate that minimizes waiting times and queue lengths in a collection of workstations. Finally, shows the effectiveness of this model through analytical methods and computer simulation applied to a series of sample cases and an application problem.

One of the most frequently occurring classes of production system can be represented by a series of productive facilities arranged so that work flows sequentially between them. These systems can be largely defined by the output characteristics of the individual productive elements of the system and the nature of the flow between them. In reality, perhaps the most straightforward series production system is where the output rate of each productive facility is not deterministic but can be described by a unimodal distribution of some sort and the flow between each facility is not mechanically paced in any way. On a macro level such a system could represent a series of major productive units feeding into each other, where overall capacity issues would be predominant. On the micro level the system describes an unpaced manual assembly line.

This paper briefly reviews the assembly line balancing techniques developed over the last 30 years. It attempts to establish the direction of research, to identify unexplored areas with potential for study and recommends future courses of action.

The purpose of this paper is to develop a computer simulation model to evaluate increasing versus decreasing mean operation times assembly line arrangement for normal and exponential distributions and the variances equal to 1 and 16.

The model was developed on the basis of a realistic case problem and applied to a six‐station assembly line. The evaluation criteria were: the minimization of the total elapsed time; the maximization of the average percentage of working time; and the minimization of the average time in the system.

The increasing mean operation times line arrangement is superior to the decreasing mean operation times line arrangement for the normal and exponential distributions and the variances equal to 1 and 16, in terms of the total elapsed time and the average percentage of the working time evaluation criteria. The decreasing mean operation times lines is marginally superior to the increasing operation times line for the normal distribution for the variances equal to 1 and 16, in terms of the average time in the system evaluation criterion. The above inference can be made for the exponential distribution for the variance 16, but no definitive conclusion can be made for the variance 1. Overall, the increasing mean operation times line arrangement has proven to be superior to the decreasing operation times line arrangement for both the stated distributions and variances, in terms of the important evaluation criteria.

The paper contributes to the computer simulation approach to solving assembly line problems that deal with the impact of normally and exponentially distributed operation times, with variances equal to 1 and 16, on the increasing and decreasing mean operation times assembly line arrangements.