Challenges in setting up and optimizing mass production lines

In the past few years, I had the privilege to help to set up and optimize mass production lines to fabricate advanced solar cells (sometimes completely from the ground up). Those were challenging times, but provided immeasurable learning opportunities. Apart from the long working hours to meet new product deadlines, these are some of the challenges during those exciting times:

1. Performance variation – Lab vs production

Building a mass production line for any new technology involves huge capital investment. It carries massive financial risks especially if there is no previous production data for the new technology. Unless it is a turnkey line, it is crucial for the new product to achieve the same level of performance (if not better) as the ones demonstrated in prototypes/labs, to meet the projected ROI.

However, this is a difficult task to achieve within a short period of time. First, production equipment have noticeable variation even among those that are supplied by the same vendors. To meet production throughput, they are also more efficient and larger in size than lab equipment. Hence, processing recipes are not 1-1 compatible, and are often not easily scalable just according to size. There are still a lot of process tuning, testing and optimization need to be carried out during line commissioning.

Second, it is also much easier to control the processing conditions and incoming materials quality in the lab. The lab equipment has low utilization rates due to smaller experiment sample size and a lack of automation. In contrast, the production equipment are expected to operate for 24/7. Thus, processing and equipment condition changes more drastically over time. This unpredictability causes inconsistent and poorer performance than the ones fabricated in a lab.

Additionally, it is also an art to know how to tune each tool individually to achieve the best collective performance. Frequently, different fabrication processes can affect one another. In fact, sometimes certain process needs to run below its potential to ensure better outcome from the other more important processes. As a result, at early stages, it is hard to ensure that all processes are operating optimally to produce the best products.

2. Unknown defects

With new equipment and the addition of mechanical automation, previously unknown defects will start to appear. At early stages, it takes a long time to identify the root causes of these new defects. This will further delay the first commercially viable product from the mass production line. Furthermore, undetected defects may not only affect the performance, but also the long-term reliability of the products.

3. Projection of operating costs

Of course, the new product needs to be commercially viable for a mass production line to be successful. However, as mentioned earlier, due to no previous production data, the operation cost of new production line can only be estimated. Often, the cost is overestimated to provide some safety nets. But this prevents more innovative products to be introduced into the market, as some projects may be considered as too expensive for investment.