assembly batches, the demand for different PCB types as well as the types of components at the assembly lines varies with time; as a result, the layout of the PCBA system is re-configured frequently and therefore must be determined. A closer look at the PCB assembly process reveals that the planning and scheduling of PCB assembly is usually very complicated.
For example, an unbalanced distribution of the assembly workload of a particular PCB type between the SMT machines in a line can cause loss of machine assembly capacity. If the workload that is assigned to an SMT machine is high compared to the workload of the other machines in the line, then the latter machines have to wait until this machine has completed its part of the assembly. When the number of orders increases it becomes very diffierent to achieve a good balance for each order; this results in idle times for the SMT machine. Therefore, there is a line balancing problem, this requires investigation via animated simulation to minimise the load imbalance between the machines; simulation results provide insight into such factors as the assembly capacity utilisation at each SMT machine or workstation. This is a very important factor at the machine planning level for the PCBA system. If the workload is equal for each machine in a line for a particular order, then the line is said to be perfectly balanced. The workload consists of picking and placing components and gluing to the boards.
Sequencing is another complicated issue during PCB assembly due to the requirement for the different types of PCB components for different board designs. To solve this problem, the PCBs are tracked from their entry into the system and throughout the processes by making use of bar code labelling and scanning.
Due to the lack of management structure in planning the assembly lines in the PCB industry, line balancing decisions were often left to the opertators[25.26.27]. Figure 2 shows a hierarchical planning and scheduling approach by Fokkert's[26] research that dealt with the complexity of PCB assembly lines. This relatively complete approach consists of three planning levels; department level planning, line level planning and machine level planning. However, it does not give any details as to how to implement this approach for scheduling and planning, particularly with a focus on the two latter levels of a complex PCBA system. Furthermore, the fatal weakness of this development is that the development models and the method are all based on a deterministic approach.However, the PCBA system is a typical stochastic system. Moreover, it also ignores the effects from the PCBA communication system, which plays a key role in such a highly antomated time-critical integrated system.
Therefore, the emphasis on developing a comprehensive but practical integrated approach for analysis, scheduling and design of complex systems such as the PCBA system, is essential in order to achieve cost-effective operations for a wide range of product types.