Quantifying the tradeoff in preventative maintenance

In the constant pursuit of efficiency improvement, the reliability of equipment in semiconductor wafer fabs is essential. The tools used in the fabrication process are extremely sophisticated and require an extensive preventive maintenance regime to ensure reliable production. All fab managers recognise the necessity of PM; however, scheduling it without a smart solution can be challenging, time-consuming, and the impact on productive output difficult to quantify.

To handle this complex problem, a fab may develop statistical models that try to predict unexpected tool downs. Scheduling preventive maintenance with a determined frequency will help to minimise unexpected disruptions in the future.

----> Download: White Paper "Superior Scheduling: hybrid approach boosts margin" <----

However, determining optimal maintenance frequencies is not an easy task, and numerous questions need to be answered in order to apply the best approach.

  • How do different frequencies of maintenance impact the fab’s ability to meet on-time-delivery?

  • How does one decide which tools to take offline together to minimise the disruption to the fab overall?

  • Is it possible to forecast whether taking one arrangement of tools offline will lead to a 4% drop in throughput for the day, but an alternative arrangement may only yield a 2% drop?

  • What is the impact of taking those same tools down at 7 am tomorrow as opposed to 5 pm two days from now?

  • What if you need the maintenance to all take place at the same time for personnel requirements? Alternatively, they may need to be staggered in 10-minute intervals for the same reason?

Smart maintenance scheduling

Answering such complex questions requires a smart solution. Flexciton's advanced scheduling technology can address all of these questions by finding the optimal schedule for your fab in any variety of forecasted conditions. Our “what-if” scenario capability allows fab managers to effortlessly trial new preventative maintenance plans according to a variety of constraints. Furthermore, rather than dictate the time that tools must be taken offline, our optimizer will find the maintenance schedule that returns the best production KPIs and thereby prescribes the optimal maintenance schedule for the factory. The fab manager merely decides suitable windows of time for each of the tools to be taken down.

This article explores three case studies where we illustrate Flexciton’s maintenance scheduling capability with varying degrees of scheduling complexity and noting the impact on high-level production KPIs.

The case studies are structured as follows:

  1. The status quo — In the first case, we optimize a production schedule with fixed maintenance timings prescribed to all start at the same time for a given toolset. Production is scheduled using a heuristic-based dispatch system emulating that found in many fabs.

  2. Optimizing production around fixed maintenance — We use Flexciton’s advanced optimizer to perform the production scheduling.

  3. Simultaneously optimizing production and maintenance — Finally, we allow the Flexciton advanced optimizer to schedule both production and maintenance timings. The window to flex the maintenance timings is chosen to be a 90-minute addition to the original timing provided in (1).

The following Gantt chart (Figure 1) shows a snapshot of the 300 lots scheduled in these small toolsets over the course of twelve hours. Each lot can only go to a certain number of tools within that toolset where the toolset is identifiable by the tool’s prefix. Each lot is assigned a priority. We optimize for the total cycle time of the lots weighted by their priority. The maintenance periods (shown in striped orange) are of varying duration and are randomly assigned to tools to take place at a specific fixed time somewhere in the twelve-hour schedule.

In Case (1), we compare the logic of scheduling, given these fixed maintenance timings, with a heuristic dispatcher against the Flexciton optimiser.

Figure 1: Maintenance starts chosen by the fab manager and production scheduled by heuristic rules.

*Not all tools shown.

Here we can see that on the tool “XZMW/097”, the dispatch system struggled to “look ahead” and dispatch effectively when given obstacles such as upcoming downtime just after 02:00. It would be better to dispatch a short processing lot in the meantime. An even more ideal schedule can flexibly move downtime around to maintain consistent, predictable throughput across the schedule.

So what if the scheduler is allowed to prescribe the timings that it finds optimal? The following Gantt chart is from Case (3), where the optimiser is free to plan the maintenance at any time within a 90-minute window.

Figure 2: Maintenance starts allowed to flex within a 90-minute window and scheduled by Flexciton optimizer.

A flexible maintenance approach makes a difference.

To present the results, we are using queue time as a KPI. As shown in the table below, evidently, optimisation outperforms a simple dispatch heuristic though clearly, this is a small, illustrative problem.

Table 1: Total queue time in hours weighted by lot priority

As we’ve demonstrated throughout this article, the driver of maintenance planning must be the fab's top-line production KPIs. In order to effectively quantify the impact that maintenance has on the production schedule, fab managers need a smart scheduling solution that can consider these multidimensional tradeoffs simultaneously. The Flexciton optimizer allows easy scenario testing and exploration in order to estimate this benefit. It empowers fab managers to answer questions such as “Am I going to critically bottleneck this tool-group in 3 days’ time by continuing to send it WIP when it requires maintenance at that time?”.

The alternative to informed decision making is that fabs are scheduling their maintenance blind and will ultimately pay the price through unpredictable cycle time, unsatisfactory throughput and unnecessary tool downtime. The idea is that by moving to smart scheduling, it is much easier to get an accurate prediction of the impact that modifying a downtime schedule will save in terms of top-level KPIs.

Enjoyed reading this article? There is more on the way. Why not subscribe to our blog updates?



In the loop

Let's talk