Why, robot

While preparation of materials on the level of material cutting (for example, the cutting of EVA and backsheets) and other basic preparation processes have long been automated by the majority of large PV manufacturers, the location of cutting and method of transportation to the production line still varies.
Fully automated manufacturing lines have tended towards in-line preparation of materials, i.e. materials are automatically prepared directly on the production line, and thus used immediately in module production. The benefit of this approach is that the materials are placed in production almost as soon as they are prepared, virtually guaranteeing that time-sensitive items, such as EVA and backsheets, will be laminated within their expiry window. It also helps prevent any foreign objects, such as dirt or dust, from accumulating on raw materials, which can occur when materials are left in one place for an extended period of time. Automatic placement of raw materials also helps ensure stability and correct placement of EVA, backsheets and interconnection ribbon, all of which can cause quality issues if slightly misplaced manually.
Semi-automated manufacturing lines, however, while still usually preparing materials in an automated fashion, will tend to manually transport and place raw materials from a dedicated material preparation room to the production line. The advantage of this approach is that it provides more line flexibility; should online automatic cutting and placement machines require maintenance or fine tuning, the entire production line is disrupted, as no raw materials can be cut orloaded on to the line while the machines are being adjusted.
Furthermore, while raw material placement devices are generally able to ensure correct material placement, it is easy for machines to be over-relied upon by factory staff that do not perform regular checks of material cutting dimensions and precise material placement positions. Clean Energy Associates (CEA) has seen cases where machines had been cutting or placing raw materials outside of factory specifications for days without any staff noticing. However, a well designed fully automated line should have all the checks and controls to achieve both higher accuracy and stability.

Tabbing-stringing, layup, interconnection

Nowadays, it can generally be assumed that all tier-1 PV module manufacturers have fully automated tabbing/stringing processes as standard. Until a few years ago, manual tabbing and stringing and layup were the norm for most facilities in China, but recently Chinese automation companies have developed affordable, sufficiently reliable tabber-stringer and layup stations. As a result, the front ends of most manual lines have been replaced with tabber-stringers and all new factories are designed with the new generation of high throughput tabber-stringers with automated layup stations.
Automated layup stations can dramatically increase the accuracy of the string matrix placement, reduce cell breakage and increase productivity. However, string interconnection is a different matter entirely. The benefits of automating this process are quite straightforward: Labor costs associated with a few soldering operators can be removed from each production line, and soldering quality stability can be greatly improved. Drawbacks of automated cell string interconnection typically come up front: Cell strings must be placed with extreme precision to ensure that automatic interconnection soldering is performed accurately. Moreover, the automation design must be flexible and allow for product changes, which is not an easy task. The initial costs of setting up such a high quality system are steep, but once the system is in place, it can provide significant stability and labor saving benefits.
Many manufacturers, especially those in lower labor cost markets, have chosen to eschew automatic interconnection because of the cost and complexity involved, choosing rather to have workers manually shift string positions at the layup process in order to ensure correct string-string distances. Once correct string position has been ensured, manual soldering of interconnection joints can be performed either by the same workers, or by others at the next station.
The benefits of both interconnection systems are clear. Automatic interconnection saves on labor, increases manufacturing stability, and is a generally impressive system when correctly implemented. On the other hand, while the improvements in stability are notable, well-trained workers at cell string interconnection stations can achieve acceptable levels of stability when performing interconnection manually, and so it is difficult to argue with the rationale of those manufacturers in regions where labor cost savings are not sufficient to justify investment in an automatic interconnection system, and thus choose to perform interconnection in a manual manner.

EL testing

Automatic EL testing software has been available for some time now, but adoption among manufacturers has generally been low. The benefits of automatically sorting laminates according to set EL defect criteria are large. However, there are no labor savings since an operator is always needed. Theoretically, a defect recognition algorithm is far more objective and stable than an operator’s judgment. It can thus assure that all modules that are unable to meet the quality criteria are removed and sent for rework; in other words, elimination of operator error.
In practice, however, the amount of false positives produced by defect recognition algorithms processing EL images of polycrystalline cells do not allow for the 100% automation of EL testing. These defect misidentifications arise from crystal defects inherent in polycrystalline wafers, which can appear as electrically inactive areas to the software, as well as darkened crystal boundaries that can have similar appearances to microcracks. These issues are significantly less when analyzing monocrystalline cells due to their naturally cleaner appearance on EL images, which allows defects to stand out far more clearly than with multicrystalline cells. Even so, misidentifications are still possible, and the benefits of automatic EL imaging for exclusively mono production lines only outweigh the drawbacks in high labor cost regions.
Some manufacturers have also taken a hybrid method to automatic EL imaging: employing automatic defect detection software as an aid for workers who still manually screen the suspect cells of every EL image for false positives. Regardless of the method used for EL testing, in CEA’s xAdvertisementexperience, high resolution cameras, high quality displays and clear EL criteria on hand for workers to consult are still the best methods.

Trimming

Trimming is a process that requires fairly high precision automation. Poor trimming can result in excess EVA/backsheet on the side of the laminate, or jagged laminate edges, which negatively affect the proper sealing. Knife movement must be smooth and uniform when performing cuts, or else trimming quality will not be equal across the edge of the laminate. The same defects can be found in manual trimming, but a skilled operator performs the quality check and corrects the deviations. CEA has observed automated trimming stations with persistent quality problems caused by poor design of the machine; at times, these defects went either completely unobserved, or were deliberately passed in order to avoid disruption of production.
Many manufacturers still use manual trimming, achieving relatively stable quality. The simplicity of the process makes automation a less compelling option, especially in locations with low cost labor. However, a well-tuned, high precision trimming station, fully integrated, is the best possible choice.

Sealing, framing and curing

Automation is of paramount importance in properly sealing, framing and curing the solar module in order to ensure its longevity. The advantage of automation in dispensing the correct amount of sealant, applying the right pressure to the frame and curing the module at the right conditions in the right time is compelling. Manually applied sealant amounts can vary greatly, either causing excessive soiling when too much sealant is applied, or insufficient module sealing when too little is applied. Framing is also a delicate procedure that can result in warped modules if pressure is not correct or uniform throughout. Automated curing can ensure the application of the first-in-first-out principle and decrease the stress applied to modules by manual handling. However, as is always the case with automated processes, a carefully designed and adhered-to list of checks must be performed at all times.

Sorting and packaging

Module sorting is an area where automation improvements are the undisputed king. The process of sorting solar modules begins when Pmax values are obtained at the sun simulator, continues through module labelling, and ends as the modules are sorted according to their power classes.
Highly automated factories will generally link all of these processes together automatically through their Manufacturing Execution System (MES).
Factories that are not fully automated in this area will generally perform some of the intermediate production steps manually. For example, some factories may have power class labels pre-printed based on expected module power classes, and require workers to manually stick these labels on modules coming off the production line.
Other factories may automatically print labels based on IV data, but still require workers to stick these labels directly on to modules. With regards to final product sorting, less automated factories will require manual worker input in order to correctly direct robots for final pallet sorting according to different power classes.
A final area that has received limited automation enthusiasm is the packaging process. Automatic packaging systems generally take up a large amount of floor space, still require manual transport of module pallets to the packing machine itself, and take an equal or greater amount of time than packaging manually.
Moderate benefits for automatic packaging do exist, the chief benefit being that all movement during the packaging process is performed with controlled and regulated amounts of force exerted on the modules.

Conclusion

As is the case with many other industries, costs come down, labor prices go up, competition tightens, quality expectations rise, and so automation gains ground on manual or semi-automated processes.
A seamlessly integrated assembly line – with minimal manual handling and in-line quality controls feeding to an MES – is the ideal choice. However, semi-automatic lines with manual steps under tight quality control can still output higher quality modules than poorly designed automated lines with quality control gaps.
Eventually, what makes a quality module, besides its design, the materials from which it is made, and the manufacturing processes it has undergone, is the tightness of the quality control checks observed during its manufacturing process. Automation is ultimately a more efficient way to control these quality checks, and to the extent that it is able to accomplish this tighter control, will result in higher product quality.