A GA-based solution approach for balancing printed circuit board assembly lines

Printed circuit board (PCB) assembly lines consist of a number of different machines for mounting electronic components onto PCBs. While high-speed placement machines are employed to assemble standard components, so-called fine-pitch placement machines are used to mount complex electronic components with high precision and by use of specific nozzles. In this paper, we investigate a typical mass production environment where a single type of PCB is assembled in a line comprising high-speed as well as high-precision placement machines. The PCB assembly line balancing problem consists of assigning component feeders, each holding a specific electronic component type, and the corresponding placement operations to machines in the line so as to minimize the assembly cycle time. To solve this problem, a two-stage solution procedure based on genetic algorithm (GA) is proposed. In the first stage, component feeders are assigned to the placement machines with the objective of balancing the workload within the assembly line. A number of candidate solutions are then transmitted to the second stage, where specific machine optimization algorithms are applied to determine the feeder-slot assignment in the component magazine of the machines and the placement sequence of the various components. As a result, fine-tuned placement operation times are achieved which reflect the individual operation mode and the actual component setup of the placement machines. Finally, from the candidate solutions the one which minimizes the actual PCB assembly time is selected.     

Development of group setup strategies for makespan minimisation in PCB assembly

Balancing setup effort and actual production time is an important issue in medium-variety, medium-volume printed circuit board (PCB) assembly. By grouping batches of similar PCB types to be processed with a common machine setup, the total time required for setting up the component feeders in the magazine of the placement machine can be reduced. On the other hand, the assembly times per PCB increase, since it is no longer possible to optimise the feeder locations in the magazine for each PCB type individually. In order to balance the savings in setup time and the increase in assembly time and to minimise the global makespan, we propose efficient grouping procedures which include machine-specific algorithms for fine-tuning the machine operations for a group of PCBs. Hence, the actual placement times are reflected much more accurately, compared with conventional grouping procedures which merely analyse the component similarity between pairs of PCBs. Two different agglomerative clustering techniques are proposed. One is based on average linkage clustering, the other on a novel hierarchical clustering approach using an inclusion tree representation of the PCB types. Both take the limited capacity of the component magazine into account. We demonstrate the effectiveness of our approach in an extensive numerical investigation of a single-gantry collect-and-place machine equipped with a rotary placement head and an interchangeable feeder trolley. Compared to conventional methodologies, the proposed group setup strategies reduce the global makespan for a given number of batches significantly.     

A MCVRP-based model for PCB assembly optimisation on the beam-type placement machine

The beam-type placement machine is capable of picking up multiple components simultaneously from the feeders in printed circuit board (PCB) assembly. Simultaneous pickup occurs only if the heads in the beam are aligned with the feeders and the nozzle-types on these heads match with the component-types on the feeders. In order to minimise the assembly cycle time, the optimisation problem is decomposed into two sub-problems, the pickup combination and sequencing problem, and the placement cluster and sequencing problem. These two sub-problems are simultaneously solved by the proposed hybrid genetic algorithm (HGA). The pickup combination and sequencing problem is similar to the popular multi-compartment vehicle routing problem (MCVRP); a genetic algorithm (GA) for the MCVRP is therefore modified and applied to solving the pickup combination and sequencing problem. A greedy heuristic algorithm is used to solve the placement cluster and sequencing problem. The numerical experiments reveal that the HGA outperforms the algorithms proposed by previous papers.     

A note on ‘balancing printed circuit board assembly line systems’

In a recent paper of (Lapierre, S.D., Debargis, L. and Soumis, F., Balancing printed circuit board assembly line systems. Int. J. Prod. Res., 2000, 38, 3899–3911.), the authors considered the balancing of a PCB assembly line consisting of several pick-and-place machines. The authors claimed that, for the particular machine type they considered, the component placement time is independent of the placement sequence of the components and then concentrated on allocating component types to machines and configuring the feeders on each machine to balance the line. We show that, the placement time is actually dependent on the placement sequence and thus, it needs to be accounted for if a more accurate line balance is looked for.     

Workload planning in small lot printed circuit board assembly

The problem of workload planning in small lot printed circuit board (PCB) assembly concerns the determination of the daily mix of production orders to be released into the production system. When switching from one production order (board type) to another, a considerable set-up time is incurred based on the number of component feeders to be replaced in the component magazine of the assembly machines. To support the order-mix decision faced by a major electronics manufacturer, two versions of a linear programming model are developed. The models differ primarily in their degree of aggregation and their computational effort. In order to reduce the aggregational error incurred, a fuzzy approach is developed to estimate the number of component set-ups at automatic SMD placement machines. Our numerical investigation reveals that sufficiently accurate solutions may be obtained from a highly aggregate fuzzy LP-model and this is achieved with considerably less computational effort than with a more detailed LP-model. We also demonstrate the potential suitability of the fuzzy LP-model for implementation within an interactive decision support system.     

Estimating printed circuit board assembly times using neural networks

Several production planning tasks in the printed circuit board (PCB) assembly industry involve the estimation of the component placement times for different PCB types and placement machines. This kind of task may be, for example, the scheduling of jobs or line balancing for single or multiple jobs. The simplest approach to time estimation is to let the production time be a linear function of the number of components to be placed. To achieve more accurate results, the model should include more parameters (e.g. the number of different component types, the number of different component shapes, the dimensions of the PCBs, etc.). In this study we train multilayer neural networks to approximate the assembly times of two different types of assembly machines based on several parameter combinations. It turns out that conventional learning methods are prone to overfitting when the number of hidden units of the network is large in relation to the number of training cases. To avoid this and complicated training and testing, we use Bayesian regularisation to achieve efficient learning and good accuracy automatically.     

An MILP model and clustering heuristics for LED assembly optimisation on high-speed hybrid pick-and-place machines

This paper deals with a scheduling optimisation problem arising in printed circuit board (PCB) assembly. In one class of PCB assembly, light-emitting diodes are to be assembled into the placement locations on PCBs by a machine with multiple pick-and-place heads. The scheduling optimisation problem is to determine the assembly sequence of placement locations and the assignment of pick-and-place heads for locations so as to minimise the assembly time. We formulate it as a mixed integer linear programming model. To solve the problem efficiently, we classify the PCBs into two types. For the first type of PCBs, on which the locations are linearly arranged, a constructive heuristic is proposed based on the analysis of the best next location after a location is assembled. For the second type of PCBs, on which the locations are circularly arranged, a heuristic based on clustering strategy and path relinking method is proposed. Computational experiments show that the solutions obtained by the two heuristics make 2.32 and 6.82% improvements averagely for the PCBs with linearly and circularly arranged locations, respectively, as compared to the solutions used in real production, and they are also better than those obtained by a hybrid genetic algorithm.     

Balancing printed circuit board assembly line systems

This paper considers the placement of components onto printed circuit boards (PCBs) using surface mount technology. Multiple automatic placement machines, a variety of PCB types and a large volume for each PCB type characterize the environment studied. The problem addressed is that of allocating and arranging the components on several placement machines, organized into one or several assembly lines, while considering a different assembly time if components are located at different feeder locations. The one assembly line problem is equivalent to balancing a multi-model assembly line where models are assembled in small batches without component rearrangement between model changes. The objective is tominimize the weighted sum of each assembly PCBcycle time, which is defined as the maximum time a PCB has to spend on each machine. We solve this problem with Lagrangian relaxation techniques. Industrial case study results are presented. We also compare the global performance of five placement machines if they are organized as a single assembly line or broken down into two or more assembly lines.     

A parallel genetic algorithm for dynamic cell formation in cellular manufacturing systems

Instead of using expensive multiprocessor supercomputers, parallel computing can be implemented on a cluster of inexpensive personal computers. Commercial accesses to high performance parallel computing are also available on the pay-per-use basis. However, literature on the use of parallel computing in production research is limited. In this paper, we present a dynamic cell formation problem in manufacturing systems solved by a parallel genetic algorithm approach. This method improves our previous work on the use of sequential genetic algorithm (GA). Six parallel GAs for the dynamic cell formation problem were developed and tested. The parallel GAs are all based on the island model using migration of individuals but are different in their connection topologies. The performance of the parallel GA approach was evaluated against a sequential GA as well as the off-shelf optimization software. The results are very encouraging. The considered dynamic manufacturing cell formation problem incorporates several design factors. They include dynamic cell configuration, alternative routings, sequence of operations, multiple units of identical machines, machine capacity, workload balancing, production cost and other practical constraints.     

Component allocation and job sequencing for two series-connected SMD placement machines

The problem of component allocation and sequencing PCB assembly jobs for two series-connected SMD placement machines is addressed. With the allocation of components, a balancing of the workload per PCB for the two machines should be achieved. Furthermore, when switching from one type of PCB to another, setup times must be considered. Two heuristic procedures are developed and tested in extensive numeric analyses of a number of realistic case models. The problem examined and the database used represent a concrete case of an industrial application in which very diversified small jobs are to be processed.     

Simulation-based performance analysis and optimization of electronics assembly equipment

Due to the high degree of automation in modern printed circuit board (PCB) assembly, production planning and scheduling in this industry relies heavily on the accuracy of the underlying models of the automated machinery. In this paper, the derivation of such mathematical models is exemplified for one of the most wide-spread machine types (the so-called chip shooter). Moreover, a simulation system for practical use in the electronics assembly is presented which incorporates various types of assembly machines. It serves as a tactical production planning aid as well as a test field for the analysis of the kinematic processes of PCB assembly machines. Such simulation, tailored to the specifics of a production environment, gives reliable estimates of the achievable production volume. Considerable deviations between the performance figures provided by the machine supplier and the actual result can thus be explained, and potential assembly system configurations be compared by drawing on reliable measures. Since the equipment is modelled at a level which takes precise account of its specific kinematics, a detailed analysis of the operations of the individual machine can also be performed revealing potential bottlenecks in its design. These insights are used to optimize machine design and control. Methods based on graph theory as well as on modern numerical search algorithms have been developed for the latter purpose. They lead to considerable performance improvements.     

Balancing and scheduling of surface mount technology lines

A mixed-integer programming approach to simultaneous or sequential balancing and scheduling of surface mount technology (SMT) lines for printed wiring board (PWB) assembly is presented. The SMT line consists of several processing stages in series separated by finite intermediate buffers, where each stage has one or more identical parallel machines. In the line, different types of PWBs are assembled using various types of electronic components. The components are assigned to feeder slots of a feeder carrier at each placement station. Different types of components occupy a different number of feeder slots. The total number of slots available at each station was limited. The problem objective was to determine an assignment of components to feeder slots at each placement station and to determine an assembly schedule for a mix of board types to complete the boards in minimum time. Numerical examples and some computational results are presented to illustrate applications of the proposed approach.     

An MILP model and a hybrid evolutionary algorithm for integrated operation optimisation of multi-head surface mounting machines in PCB assembly

This paper focuses on an operation optimisation problem for a class of multi-head surface mounting machines in printed circuit board assembly lines. The problem involves five interrelated sub-problems: assigning nozzle types as well as components to heads, assigning feeders to slots and determining component pickup and placement sequences. According to the depth of making decisions, the sub-problems are first classified into two layers. Based on the classification, a two-stage mixed-integer linear programming (MILP) is developed to describe it and a two-stage problem-solving frame with a hybrid evolutionary algorithm (HEA) is proposed. In the first stage, a constructive heuristic is developed to determine the set of nozzle types assigned to each head and the total number of assembly cycles; in the second stage, constructive heuristics, an evolutionary algorithm with two evolutionary operators and a tabu search (TS) with multiple neighbourhoods are combined to solve all the sub-problems simultaneously, where the results obtained in the first stage are taken as constraints. Computational experiments show that the HEA can obtain good near-optimal solutions for small size instances when compared with an optimal solver, Cplex, and can provide better results when compared with a TS and an EA for actual instances.     

Petri nets and genetic algorithms for complex manufacturing systems scheduling

In this paper we propose the GAPN (genetic algorithms and Petri nets) approach, which combines the modelling power of Petri nets with the optimisation capability of genetic algorithms (GAs) for manufacturing systems scheduling. This approach uses both Petri nets to formulate the scheduling problem and GAs for scheduling. Its primary advantage is its ability to model a wide variety of manufacturing systems with no modifications either in the net structure or in the chromosomal representation. In this paper we tested the performance on both classical scheduling problems and on a real life setting of a manufacturer of car seat covers. In particular, such a manufacturing system involves features such as complex project-like routings, assembly operations, and workstations with unrelated parallel machines. The implementation of the algorithm at the company is also discussed. Experiments show the validity of the proposed approach.     

Precedence constrained TSP arising in printed circuit board assembly

Component placement sequencing is a challenging problem that arises in automated assembly of printed circuit boards. While for some placement machines all placement sequences are acceptable, in other cases some sequences are not allowed because of the shape of the placement head. In such cases, while the head moves down to perform a placement, it might damage a previously placed component, and the problem of determining a minimum cost and at the same time acceptable sequence leads to a Precedence Constrained Travelling Salesman Problem formulation. In this study, a solution procedure to such a formulation is developed and its implementation in a real PCB assembly environment is discussed.     

Analysis of partial setup strategies for solving the operational planning problem in parallel machine electronic assembly systems

In this paper, we analyse the operational planning problem in an electronic assembly system with multiple placement machines operating in parallel. The partial setup strategy is proposed as an effective methodology for this problem. This strategy attempts to determine the balance between processing time and changeover time during system operation. The Tour primary issues are determining the assignment of products to machines, the sequence of products on each machine, the assignment of components to feeder locations for each product, and the component placement sequence for each product. Four solution procedures for unique and partial setup strategies are developed and tested. The computation results show that for low to medium volume high mix environments, the partial setup procedures offer significant improvement over the unique setup strategies. The results also show that no single fixed strategy dominates in all scenarios, and therefore, an adaptable procedure that will choose the best solution for each set of requirements is needed.     

Mechanism to minimise the assembly time with feeder assignment for a multi-headed gantry and high-speed SMT machine

This paper proposes the development of a mechanism to minimise the assembly time for a multi-headed gantry and high-speed surface mounting technology machine by determining the component assignment to feeder slots. Since a gantry moves long distances in order to pick components, place them on the board and then return them to the feeder slots, we classified the overall assembly time according to the different movements of a gantry. The overall assembly time is then minimised by presenting a new heuristic for the feeder assignment, consisting of Nearest Component Allocation (NCA) and Globally Updated Assignment (GUA). NCA contains information about how each component type locates closely to others on the board. Using the solution from NCA, the component distance function calculates the most representative distance between component types. Then, GUA is applied to improve the NCA solution. The experiments consist of several printed circuit boards with numbers of component types and points to be placed. Highlights of this paper are that: a classification of the gantry movements is proposed based on the average speed; a heuristic NCA-GUA for feeder assignment is developed by considering the placements on the board; the computational time is greatly reduced by NCA-GUA without degrading the solution quality; and a decision process for nozzle assignment is proposed to minimise the overall assembly time. The results show how NCA and GUA affect the final results, and how this mechanism leads to better performance than a genetic algorithms or 2-opt swap search. This comparison proves that our method provides competitive and effective solutions in terms of minimising the overall assembly time.     

AUTOMATED PROCESS PLANNING FOR PRINTED CIRCUIT CARD ASSEMBLY

Printed circuit card assemblies form the core of a vast array of contemporary manufactured products. The technologies for assembling printed circuit cards require a hierarchy of complex decisions for grouping card types and processes, staging components at assembly machines, arranging feeders, and sequencing placement operations. This paper is motivated by the largely unmet needs of industrial process planners for computer aids. Our objective is to develop a framework for the printed circuit card assembly process planning problem and to assess the current state of the research on appropriate models and solution methods. We first provide an overview of the essential elements of printed circuit card terminology, assembly technologies, and assembly system operations. Then we propose a decision hierarchy, survey the published literature, and identify needs for future research.     

Developing a method to generate semi-distributed layouts by genetic algorithm

Recent research in industries shows that existing layout configurations do not satisfy the needs of multi-product enterprises in turbulent environments but within new layout strategies, distributed layouts have deserved more attention in most manufacturing environments and have a promising potential to cope with demand disturbances. This study is an attempt to design weighted distributed layouts via considering machine independent capabilities by a resource elements (REs) approach, which has caused generation of a new type of distributed layout named semi-distributed layout. REs are used to define processing requirements of parts and processing capabilities of machines. Another contribution of this paper is applying genetic algorithms (GAs) to distribute REs to find the optimal assignment of machines to available locations in such a way the travelled distances of parts are minimised and the accessibility of them to the required machines are maximised. The methodology of this paper is illustrated using a two-phase procedure. First, all machining facilities are divided into a set of REs based on their capabilities and second, the weighted connections among REs are considered to distribute them over the floor through implementing the developed GA. To evaluate the methodology, the proposed algorithm is tested with three illustrative examples obtained from the literature, in which two of them are comparable with outputs of simulated annealing (SA). The comparison between the outputs of the GA and the SA on the same cases presents that for large size problems, the GA significantly outperforms the SA.     

Flexible job-shop scheduling problem under resource constraints

A flexible job-shop-scheduling problem is an extension of classical job-shop problems that permit an operation of each job to be processed by more than one machine. The research methodology is to assign operations to machines (assignment) and determine the processing order of jobs on machines (sequencing) such that the system objectives can be optimized. This problem can explore very well the common nature of many real manufacturing environments under resource constraints. A genetic algorithm-based approach is developed to solve the problem. Using the proposed approach, a resource-constrained operations–machines assignment problem and flexible job-shop scheduling problem can be solved iteratively. In this connection, the flexibility embedded in the flexible shop floor, which is important to today's manufacturers, can be quantified under different levels of resource availability.