Renewable energies will be the key technologies of this century. In the case of photovoltaics, i.e. the direct conversion of sun light into electrical power, the world market was growing by 25% per year over the last 5 years. This increase rate is comparable to those of information technologies and telecommunication.

VISOLAR addresses an improved quality management system in the production of poly-crystalline silicon solar cells and modules by visual inspection. Conventional quality management systems are using only electrical measurements. VISOLAR established additional trainable visual inspection tools in the standard quality management to reduce breakage rates, increase through-put, improve quality of the solar cells and photovoltaic modules based on advanced image processing, feature extraction, and pattern recognition.

The VISOLAR Inspection System

VISOLAR offers an improved quality management system in the production of poly-crystalline silicon solar cells and modules by visual inspection. The system modules are fast enough to allow the integration in modern production lines with throughput rates of 1200 solar cells per hour. This equals to an annual production capacity of about 15 MW or 25 MW photovoltaic power for one line.

The top priority issue is to reduce the breakage arising from the soldering of the cells. Micro cracks of solar cells that are invisible to the human eye can lead to breakage in subsequent handling and production steps. Those cells should be deselected before arraying the individual cells up to several hundreds in large-size photovoltaic modules. Especially for the case of transparent glass/glass modules the breakage risk becomes an important cost factor. The glass/glass technique shows an inherent higher breakage rate, especially when used in high-throughput vacuum laminator for high production rates. This risk scales with the increasing size of these modules for facade integration.

Aesthetic flaws of the crystalline cells that reduce the appearance of the photovoltaic modules are recognised by the system and can be deselected as well. This has become even more important since photovoltaic modules became a considered component in facades and other building integrated applications. The requirements coming from architects has increased the aesthetic demand on the product. Especially for the case of coloured cells, as they were used for the photovoltaic facility on top of the Bundespräsidialamt in Berlin, small variations of the desired colour obstruct the uniform appearance of the module or the facade. Aesthetic flaws in general often can not be defined precisely. The range of what is acceptable or not has to be defined by the individual manufacturer. This is achieved by training the inspection system with a set of aesthetically acceptable solar cells. Especially for delivery contracts relying on aesthetic flaws, this provides a possibility to control the quality of the solar cells for both parties, the purchaser and the vendor as well.

Deselected cells can be used in B or C class modules or in modules with lower production risk. In addition to the quality and yield improvement of this approach cost reductions are achieved from the possible higher degree of automation. Because it also covers the aesthetic classification it can be used when the inspection by the human eye is no more available.

This was demonstrated successfully for the new soldering equipment of the SOLON AG. Because of its low breakage rate the machine is highly competitive compared to existing machines on the market. At a maximum through-put rate of 700 cells/h the breakage rate is by a factor of 6 lower than the standard breakage of 2% as usually offered. The machine was designed by SOLON in co-operation with GfaI and with Jonas & Redmann as manufacturer of the machine.

The inspection system consists of various modules in a software package for

• Detection of micro cracks and v-shaped chips,
• Colour classification and,
• Detection of aesthetic flaws on the front and back side of a solarcell, including the camera, frame grabber, and illumination specification. The modules are trainable for the considered type of cell or wafer.

Eine Aktivität des HPCN Technology Transfer Node - Thuringia (TTN-T)

Projektpartner

• GFaI (Coordinator)
• SOLON AG
• Parsytec GmbH
• TraB

Weitere Informationen

Public Final Report

European Integrated Machine Vision Newsletter. Special Edition: Manufacturing. vo. 1, 2003

Film Löten

Film Legen