What are you measuring, and what are you doing with the data?
Sensors and monitors are great for tracking process improvements and alerting equipment operators if a process drifts out of spec. But they can also help with greenhouse gas (GHG) emissions reductions.
Smart manufacturing is about using technology to make manufacturing more efficient and reduce errors. In the semiconductor sector, that typically means focusing on yield. The better you can track where defects are occurring and what types of defects you see, the better informed you are to make relevant process improvements.
The same is true for meeting sustainability goals for GHG emissions, energy consumption, and other sustainability metrics. Process optimization can target productivity and sustainability goals simultaneously. Changes that reduce energy consumption, save steps, or shorten process duration without sacrificing performance win on both types of goals.
Today’s technology is sophisticated enough to optimize for multiple parameters without the need for extensive experimental testing. Sensors share data in real time. Continuous monitoring allows operators to adjust on the fly. Machine learning (ML) software can allow the equipment to monitor itself, halting a process well in advance of complete failure so that repair and replacement can happen proactively.
Water Quality at Home and in the Fab
I’ll admit that ML is not my area of expertise, but I think of an analogy to my refrigerator water filter at home. My fridge uses a simple system without a feedback loop. It keeps track of time since the last filter replacement, gives a warning to replace the filter soon, and the beeps with a “replace now” notification when six months have elapsed since the last manual reset.
The six-month mark is merely a guideline. Actual effective filter life depends on the quality of incoming water and how much water flows through the filter each month.
A smart system would sense water quality and give a “replace soon” notification when it first begins to decline and then a “replace now” notification when it declines to an unacceptable level. An ideal system would distinguish between contaminants that are unhealthy or dangerous and those that merely affect the water’s taste.
With a smart system, residents gain confidence that their water is safe. If the recommended replacement interval is longer than six months, such systems reduce the inherent waste and expense of too frequent filter replacement. If the water quality is declining faster than expected, it may be a sign that the incoming municipal water supply is substandard. In that case, you might want to install additional whole-house filters or contact the local utility and demand safer water.
This level of sophistication might be overkill for a home fridge, but it isn’t for a fab where water quality makes the difference between a process with an acceptable yield and one that’s a disaster.
Company-wide Collaboration
Optimizing for sustainability means looking at far more than water filtration efficiency and filter replacement rates. Any tool or process step that consumes energy and resources is a candidate for improved sensing, tracking, and predicting.
The first step is to measure parameters in situ. Systems that continuously monitor energy and GHG consumption, product yield, and effluents indicate where to make changes. In some cases, this can occur automatically. In other cases, engineers can use the data to inform decisions.
Examples include adjusting process temperatures to reduce demand for heating or cooling, placing features in idle mode when not needed, or using reusing heat generated in one process step to power another part of the process. Real-time data removes the guesswork in making any of these improvements.
Sustainability works best when it’s not a standalone function but is incorporated throughout the company. A network of smart sensors can make this easier by sharing data seamlessly. When sustainability champions and process development experts have access to the same data, they can collaborate to advance company-wide goals.
What Does This Mean for Your Company?
Since process gases with high global warming potential (GWP) make up the largest part of GHG emissions, any industry-wide plan to incorporate smart manufacturing needs to find ways to reduce emissions from plasma etch. If you work with process gases, replacing those with the highest GWP is probably already part of your roadmap.
Even if your company’s products are far removed from dry etch, smart manufacturing is worth considering. You can evaluate the greatest sources of GHG emissions and energy consumption in your processes. Developing in situ sensing systems with feedback loops that target these sources will improve efficiency and yield while helping you meet your sustainability goals.
