, on February 27, 2024, 11:42 AM

The Power of Inaction: Strategic Decisions in Semiconductor Manufacturing

Some time ago, I served as an Emergency Response Team (ERT) leader at a semiconductor facility in the US. I was charged with coordinating safe and effective responses to emergency situations. My team faced a range of issues – from personal injuries to gas leak detections to spills of gaseous and liquified chemicals with multi-syllabic names. At the onset of any emergency, we had three tasks to complete within the first few minutes of the call:

  1. Assess the situation
  2. Identify possible responses
  3. Implement a response

One response often went unstated but was always on the table: “Do nothing.” It’s a very counter-intuitive response, especially when risks to equipment, environment, and personnel exist. But sometimes, doing nothing is the right response.

I’ve recently considered this question in terms of the manufacturing process. How can it be possible for even the most senior floor operators to know, with confidence, exactly what lot to run given the hundreds of seen and unseen variables that exist in a manufacturing facility at any minute? And how can doing nothing when choosing lots to process actually be a benefit?

Materials that queue in front of equipment likely have specific and unique demands. One might be marked as “hot.” Another may contain a scheduled due date that results in a favorable (or unfavorable) critical ratio. Still, another may represent a product that, if shipped by this weekend, will bring in huge amounts of present and future revenue. Faced with these complex and often competing priorities, it’s not hard to imagine a floor operator facing a type of emergency when determining what to run next, especially when a process run on a given piece of equipment could take 24 hours or more to complete. That’s a lot of time to commit a tool to a process! The natural, human response at these points of decision is to pick anything and process the material so that “something is running”...a response that can result in unintended impacts to upstream and/or downstream throughput or utilization, missed due dates, bottlenecks...the adverse effects are too numerous to name.

To be clear, there are situations where “just process whatever is there” is the correct response. But have you considered not processing the material in support of a greater goal, metric, or outcome?

Here are a few situations where “doing nothing” may be the best response:

  • At equipment where the same raw material is consumed – whether one lot or six lots are processed (batch equipment)
  • For count-based maintenance plans where maintenance actions are determined by a wafer or lot count limit
  • For scheduled processing of a lot that has the absolute highest priority in the fab
  • Full batch processing that benefits near-term downstream tools
  • Processing that maximizes the utilization of time-bound raw materials (such as a chemical bath life limits)

Doing nothing forces both human and automated systems to work cooperatively toward decisions that incorporate lot context, equipment state, downstream demands, and a host of other states and statuses. Doing nothing may help improve specific KPIs such as optimization and equipment utilization. Doing nothing provides room for those last-minute changes that occur before the material is formally committed to the equipment.

When you think about it, we face decisions all the time where the most reflexive response is often the worst response – when in fact the counter-intuitive response is the best.

Here are a few examples:

  • If you find yourself caught in a riptide in the ocean, the worst thing you can do is to flail your arms and legs – the best response is to stay still and try to float.
  • When financial calamity occurs, the worst (and easiest) response is to take on more debt – the best (and most difficult) response is to establish a plan to cut spending and live on a budget.
  • In 1968, mathematician Dietrich Braess studied vehicular traffic patterns through bottleneck sections of freeways and intersections, and he mathematically proved that adding additional routes to a fixed destination (a reasonable response) increased the level of traffic, whereas reducing routes to the fixed destination (a counterintuitive response) reduced traffic and helped minimize the effects of the bottleneck. His findings became known as “Braess’s paradox”.

The decision to “do nothing” can be a profound strategy in both emergencies and manufacturing processes, yet it requires careful consideration and analysis. This is where the significance of scheduling and dispatching becomes evident. By integrating these automated systems, we can more effectively harness the power of inaction. These systems not only guide the operator towards the most efficient and effective decisions by analyzing a multitude of variables but also empower them to recognize when inaction is the most beneficial course. Embracing automated scheduling and dispatching enables a data-driven approach to decision-making, ensuring that each choice is made with the utmost clarity and insight – and instilling confidence when that choice is strategic inaction.


No comments published.
Be the first to comment.