Search results

Surface mount technology (SMT) is being increasingly used in printed circuit board (PCB) assembly. The reduced lead pitch of surface mount components coupled with their increased lead count and packing densities have made it imperative that automated placement methods be used. However, the SMT placement process is often a bottleneck in surface mount manufacturing. A reduction in placement time in SMT will enhance throughput and productivity. This paper describes the design and development of a prototype expert system based approach which identifies ‘near’ optimal placement sequences for surface mount PCBs in (almost) realtime. The software structure used integrates a knowledge based system with an optimisation module. PROLOG is the language used in this research. The system was rigorously validated and tested. Ideas for further research are also presented.

In order to handle high and rapid production demands, printed circuit board (PCB) manufacturers have employed high‐speed surface mount machines into their assembly lines. These machines have abilities of fast component placements, but provide challenges for process engineers to optimise the component placement sequences and feeder arrangements via effective planning. A computer program was developed based on operating concepts using genetic algorithm, to solve for various component placement sequencing planning of the high‐speed chipshooter. Genetic algorithms are a class of general purpose search methods based on the concepts of genetic evolution and survival of the fittest. The program provides information on component placement sequences and feeder arrangements for optimal assembly times. Initial tests have shown that the size of the parent space affects the convergence of the solutions during iterations. Finally, comparisons of results have shown improvement over those previously obtained by other researchers.

In a high‐mix middle‐volume production environment for printed circuit board (PCB) assembly, the production efficiency strongly depends not only on the tactical level of how to group PCBs but also on the operational level of how to assign feeders and determine placement sequences in the group setup strategy. The present study discusses the problem of clustering PCBs into groups in such a way that total placement and setup time can be minimized. This problem is motivated by a situation that the reduction of group setup‐time and efficiency loss of placement time should be balanced in a PCB group setup optimization. This research incorporates placement time into the PCB job grouping and presents a weighting similarity measure. To solve component‐feeder assignment and placement sequences for a family of PCBs, an efficient procedure based upon an ant colony optimization (ACO) algorithm is developed. Group setup performance is evaluated and compared under a variety of grouping algorithms. Experiments are conducted to discover situations where the consideration of efficiency loss of placement time can make a significant improvement in PCB group assembly.

Notes that, until now, to route robotics travel, most investigations have utilized the fixed coordinate of placement points and magazine of the traveling salesman problem (TSP) method to sequence the placement points after the magazine has been arbitrarily assigned. Points out that, in fact, robotics travel routing should be based on a relative coordinate because the robotics, board and magazine simultaneously move at different speeds during assembly. Consequently, the coordinates of placement point and magazine are constantly changing. In this study, a novel tabu search (TS) based approach is presented. The proposed approach can arrange the placement sequence and assign the magazine slots to yield a performance better than the conventional one. Results presented herein also demonstrate that the larger the number of placement points and/or part numbers, the better the performance.

This paper presents an investigation on the development of different pattern placement strategies in robotic palletisation of box packages in the packaging industry with practical implementations for one, two, four and five block patterns with the aim of improving the operational efficiency in robotic palletisation.

The work involves considering the gripper design and maximum number of picks and various process parameters that affect the robotic implementation of pallet patterns and develops a methodology to form different patterns for a given pallet size.

The proposed methodology represents an efficient approach for pallet pattern implementation and results in reduced number of placements required for a given number of boxes per layer and reduced time for palletisation.

The paper introduces a novel technique for pallet loading problem (PLP) considering the physical aspects and restrictions encountered when using the robot and the gripper size to generate the pattern on the pallet. Traditional solutions of PLP do not consider these aspects in pattern placements.

To highlight the differences and common features of taboo search (TS) and genetic algorithms (GA) in solving the problem of board‐type sequencing on the assembly line simultaneously with the combined problem of feeder assignment and component placement sequencing in the printed circuit board (PCB) industry.

Two metaheuristics (search techniques) are used to solve three problems associated with the PCB assembly line: TS and GA. The implemented approach is used to solve the three problems on a single pick‐and‐place sequential machine with a stationary board table and stationary feeders, and with the use of the Euclidean metric.

The achieved results show a satisfactory reduction in assembly time, when TS and GA are compared with a random solution, with a slight superiority of TS over GA. However, the program running time is longer for TS.

The hypothetical case study used shows that in real life the savings could reach an average of 6 per cent when TS is used. Slightly lower savings are possible when GA is used.

This paper provides a clear insight into how some of the problems associated with the production of PCBs can be solved simultaneously using metaheuristics such as TS and GA.

A printed circuit board (PCB) assembly consists of a circuit board and a variety of components in different shapes and sizes placed at specific locations on the board. They are used extensively in a variety of products by many industries today. These products include, for example, computer terminals, televisions, electronic instruments, and process controllers. There are many types and models of automatic machines available for placing components on the boards. The component placement operation consists of all steps necessary to remove a component from its packaging materials and mount it onto the PCB. The components are supplied either from a set of feeders, magazines or matrix trays. Because of the extreme accuracy required, automatic equipment is mandatory for all but the smallest production volumes.

This paper considers the decision problem for a minimum setup strategy of a production system arising in the assembly of printed circuit boards of different types, using a placement machine with multi‐slot feeders. We formulate the problem as a binary linear programming model, and propose a heuristic procedure to find the solution that consists of a board‐assembly sequence, an associated component loading and unloading strategy and a feeder‐assignment plan within reasonable computational effort. Computational results from solving the simulated problem instances by using the heuristic method and the mathematical model are provided and compared. The proposed heuristic procedure can be incorporated into the PCB scheduling optimization software to decrease cycle times and increase overall assembly throughput in a high‐mix, low‐volume PCB manufacturing environment.

The purpose of this paper is to propose an effective and efficient solution method for the component allocation problem (CAP) in printed circuit board (PCB) assembly, in order to achieve high‐throughput rates of the PCB assembly lines.

The investigated CAP is intertwined with the machine optimization problems for each machine in the line because the latter determine the process time of each machine. In order to solve the CAP, a solution method, which integrates a meta‐heuristic (genetic algorithm) and a regression model is proposed.

It is found that the established regression model can estimate the process time of each machine accurately and efficiently. Experimental tests show that the proposed solution method can solve the CAP both effectively and efficiently.

Although different regression models are required for different types of assembly machines, the proposed solution method can be adopted for solving the CAPs for assembly lines of any configuration, including a mixed‐vendor assembly line.

The solution method can ensure a high‐throughput rate of a PCB assembly line, and thus improve the production capacity without further investment on the expensive PCB assembly equipment.

The paper proposes an innovative solution method for the CAP in PCB assembly. The solution method integrates the meta‐heuristic method and the regression method, which has not been studied in the literature.

This paper studies a production planning problem in printed circuit board (PCB) assembly. A PCB assembly line has several non‐identical placement machines, so the placement times by different machines are various to the same type of components. Several mathematical models are formulated in order to obtain a best assignment of components to machines with the objective of minimizing the cycle time to have the best line throughput. Moreover, the data structures of the models are analyzed and compared with other similar models to search a good available algorithm. Finally, a numerical example is provided to illustrate these models and is solved by a computer package.