Reading time ( words)
By Adrian Schärli, Essemtec AG
Short-term change of priorities, rush jobs, and small production volumes are the norm in today's SMT production. In such an assembly environment, the largest costs are accrued by machine set up, set-up change, and product changeover. These costs can be minimized through set-up optimization and parallelization. It is a concept that saves costs and increases productivity. It also puts high demands on personnel, your organization, and placement machines.
Medium-size lots may lower the average productive time of a non-flexible, trimmed-for-speed placement system to less than 50% of total manufacturing time (Figure 1). The rest is lost to unproductive setup operations since standstill due to technical problems hardly can be shortened any further. A highly flexible intelligent placement system machine can extend machine runtime based on the same volume by up to 30%. In a single work shift operation, this adds 2.5 hours a day. At an estimated cost of $200/hour per line, this amounts to a typical daily profit gain of $500, or about $100,000/year.
Retrofit vs. New SetupWith a non-flexible system, the imbalance of runtime and standstill will grow much worse as the batch size gets smaller. In contrast, the runtime of a highly flexible machine featuring an intelligent set up concept will decrease only marginally, even when producing very small lots.
Setup changes typically go together with a change of printed circuit board (PCB) format and at least some of the components to be placed; components are the largest changeover expenditure. An important consideration at setup is that manufacturing time is strongly correlated with the spatial distribution of the various components on the component placement machine. Oft-requested components should be loaded close to the PCB to minimize placement head travel. An IC that is used only once per board may be located farther away.
At this point, an operator might ask if it is really worth the trouble to position all those components at an optimum for the follow-on job. Wouldn't the shifting of the feeders waste more time than can possibly be saved through placement optimization?
Simulation of Different Changeover StrategiesThe answer to this question hinges on a simulation of the follow-on job based on the actual set up situation.
The considerations, for two models, include manufacturing time of the follow-on job after optimizing the position of all components (complete new setup), manufacturing time for the follow-on job when current component positions are left intact. Only additionally needed components are retrofitted.
Looking at the results, we can compare the total time for changeover and manufacturing in Table 1.
In this example, 37 minutes would be gained if only new components were retrofitted and optimized component positioning is declined.
Minimizing Set-up CostMany productions work with a fixed set up and large numbers of fixed standard components such as resistors and capacitors. Any simulation must account for this. Another consideration is how much stock is remaining on the various rolls, sticks, or pallets used.
Such a simulation cannot be carried out weeks ahead of the production start; there are too many changes possible in the production plan. Instead, the simulation must be based on the actual situation. It is best to carry one out directly in the ongoing production. Figure 2 shows a possible flow diagram.
Figure 2: Simulation software can determine the most efficient procedure (retrofit or new setup) at any given moment.
The more feeders and locations available, the more components can be set up at a time. This enables fixed set up of standard components and also complete set up of multiple jobs (bundle). It is significantly more efficient to set up an entire daily production at once.
Feeder Changeover During ProductionEvery minute a chip shooter or pick-and-place placement machine is standing still, it wastes money and does not deliver value. Thus, feeders must be accessible for set-up or take-out regardless of the running operation without stopping the machine. Proper marking or blocking is needed to ensure that none of the feeders currently in use can be removed.
Flexible manufacturing requires feeders that are set up and programmed off-line. To ensure placement quality, the feeders are intelligent. The machine must recognize feeder types and their positions automatically. The simplest and most effective way of doing this is numbering the feeders with unambiguous IDs. The system, upon recognizing an ID number, fetches component type, marking and remaining stock for the actual setup from the data bank. The system then reads the used-up stock data back. This way, demand planning is always based on an updated list of stock. Another positive side effect is that series errors due to falsely set-up components are eliminated.
Placement System RequirementsThe following summarizes a placement system's most important features, which contribute significantly to cost reduction in a flexible manufacturing:1. Large capacity for feeders the more, the better (Figure 3).2. Feeders are exchangeable in operation.3. Intelligent feeders which have positions and content recognized automatically. 4. Feeders can be set up offline and programmed via barcode reader.5. A simulation and optimization tool to calculate manufacturing times in various setup environments. This tool should accommodate various situations including actual machine setup and fixed setups, actual component stock remaining on rolls, and a combination of several jobs (bundle setup). It also needs to take into account placement times for different components, travel from feeder to placement position, and PCB transport times.
A placement machine with these features will optimize production capacity at any time and in any situation. Wait times, changeover times, and standstill will be minimized, maximizing productivity.
ConclusionWho are the users of intelligent placement machines? Often they are mid-size enterprises and manufacturing services producing complex systems in medium-size lots. They must accommodate frequent product changes as well as prototype manufacture on short notice. They expect continuous manufacturing through minimum standstill and avoidance of series errors. Comprehensive tracking of component stocks ensures that jobs are commenced only if all components are on hand in sufficient quantities. In most cases, these intelligent systems realize short-term capital amortization.