Lesson 2: Low or High Automation

Below the information boxes, you will find the text version of the video and further downloads. You will also see the title of the next lesson and when it will be available:

The next lesson will be available about 3 days after this lesson. When the lesson is available, you will receive another email from our system with a link to the video.

In addition to the video, we have prepared a PDF version of the board in this video, which can be viewed and downloaded.

You can automate processes for two main reasons:

1. Quality advantages
2. Lower costs

Below we offer some examples of these two reasons.

It makes sense to automate all the processes that are critical for the quality of your output. You can automate not only the production process itself (stringing, layup and so on) but also the quality check. There are many quality systems that verify quality automatically.

There is no guarantee that automating all parts of your production line will lead to a much higher quality product!

Even with 100% manually produced panels, you can still produce a solar module of very good quality.

But the work of a human is not always consistent! Some days their work may be better and some days it may be worse.

A machine will produce consistent quality all the time: if you adjust the machine well, it will make perfect solar modules and each one will be equally as good as another. It will be much easier to improve your output quality if you have a highly automated solar module production line.

This is why in the long term, a more automated production line will make panels of higher quality than a manual production line.

If we consider the process of solar module production, some parts are especially critical for determining the quality of a solar module, while other parts are less important and less critical to automate. Below we list some critical areas of production: 

Cell soldering:

Solar cells are very thin and can break easily! If the solar cell is broken in its microstructure (which you can’t see with your eyes alone), then the power of the solar module will decrease over time much faster than a perfectly soldered cell. 

This is why it makes sense to automate this step with an automated stringer. 

Consider automating this process first before thinking about any other automation. 

We recommend making stringing fully automated even with the smallest line. 

Layup of the strings:

The next production process involves taking the strings of solder cells onto the glass. 

This process is also very risky to do by hand because touching the cells can crack them. Also, if you use a manual gripper and move too quickly, this can cause damage!

The automated solution is a layup station with automated grippers or robots that place the strings very softly on the panel. In addition to the technical advantage of this station, you will get a much better result in the "optical quality." This means that all the strings will be 100% placed in the same way (the same space in between, and completely parallel strings). This will make a better impression on your customers. If your line is about 20 MW or bigger, you should definitely make this process fully automated!

Soldering of the sides (busing):

After the strings have been placed on the panel, they must be connected to produce power and transport it to the junction box. It is very important that the soldering connections are 100% perfect to avoid hot spots and losses. 

Of course, the safest way to solder the sides is a fully automated soldering process - but on the other hand, this is very expensive. Because of the costs, we do not recommend an automated soldering system for the sides for lines below 100 MW. In all other lines, you should keep this process manual. 

Lamination process:

The lamination process itself is always automated and cannot be done by hand. But there are differences in the way you load and unload the laminator.

The loading of the panels is especially critical, because:

• every solar module must undergo the same production process. If one solar panel is heated a bit longer than another, the quality will not be the same!
• when you load the panel, it is not a package but all separate parts. Any handling of these components can damage the panel and the cells.

With a fully automated process for loading the laminates, you can ensure that every lamination is of the same quality. We recommend making the laminator fully automated if you have a 20 MW line or more!

Of course, there are many other processes where it makes sense to automate, but these four steps are the most important areas.

For many (but not all) quality checks, you can complete them manually. But you need to think about the long-term reliability of each quality check.

A human can be very good at a quality check, but while sometimes they may be 100% accurate, at other times they may be tired or ill and therefore less efficient.

The result is a quality check that is not always consistent and therefore an inconsistent quality of the product. The goal for every production line should be consistent quality across all processes and at all times.

This is why it makes sense to automate quality checks. For example:

• Optical cell tester in the stringer: ensures that there are no damaged cells in the strings.
• String tester: checks that every soldered string produces enough power.
• EL-Tester: shows whether there are (hidden) micro cracks in the cells that will cause problems in the future.
• Hi-Pot Tester: determines if there are any leakage currents in the panel isolation.
• Flasher: ensures the panel has the power that you claim! The flasher is one of the most critical parts of production, and you really should find a highly accurate one.

Another aspect of quality in solar panels is their visual appearance!

Even if a panel is 100% electrically sound, you may have problems selling it for the normal price if its appearance has problems like:

Unparallel strings
Varied cell distance
A frame that is not precisely 90°
A bent frame
A label that is not consistently in the same place

And so on! This has nothing to do with electrical quality but rather the visual appearance of your products. 

Automation can lead to higher quality products, or it can replace employees and therefore save money. It is simple to determine whether it makes sense to automate a process to reduce costs: you can easily calculate how many workers can be replaced by a robot or automation. Then compare the salary of the worker(s) with the costs of the automation to find out how long it will take to make back the money you would spend on the automation.

The results of these calculations depend on the salaries of the workers. For this reason, an automation in a country with low salaries makes less sense than in one with high salaries.

Next, let’s look at some examples of places where automation can replace human workers.

Example 1: Cell soldering:

Cell soldering can be done manually, but it requires a lot of employees. To give you an idea: a new stringer can make about 1.200 cells per hour, which is about 40 MW. To operate this stringer, you need 1 operator per shift, so 3 people across 3 shifts.

To make the same number of cells with manual operators, you would need 12 people per shift, so 36 in 3 shifts.

As a result, if you make 40 MW of automated stringing instead of manual stringing, you save around 30-33 employees. This is substantial, and it is one of the reasons why even in developing countries with lower labour costs, the stringing will be done automatically at a size of 20-40 MW or more.

However, it might make sense to use manual stringing if your production line produces very few MW, such as 2/3/4/5.

Summary: It makes sense to automate cell soldering not only to improve quality but also to save money.

Example 2: Trimming/Taping:

Trimming and taping can be done either manually or with full automation. There is no advantage in quality due to automation, but of course you can reduce the number of employees required and therefore save money.

One employee can make trim and tape for about 50 MW. A single robot can also do both steps and has a capacity of up to 100 MW.

Let’s compare these two methods:

100 MW manual trim / tape -> two workers per shift; 6 workers in 3 shifts

100 MW robot -> one robot

If you want to determine the return on investment over 3 years, you can calculate:

6 employees + 25% (to account for holidays, for example) = 7,5

So over 3 years, you must pay 22,5 in salaries.

The robot will cost a minimum of 100.000 € more than the manual solution.

Result:

If the monthly salary of one worker is 370 € or more AND you can ensure that in the first 3 years you will operate 100% of the time, then the manual route might make sense.

Summary:

In my opinion, even with a 100 MW line, it makes no sense to make the trim / tape with a robot. You will lose cash and flexibility. Also, the risk is high if anything changes in your production line and causes you to have to invest again in a new robot. So trim / tape is not an area of production where I would spend any money on automation.

This type of calculation can be done for any other step of the production line in the same way as above.