Machine vision optimizes thin-film solar production

Six industrial cameras,combined with vision software, are maintaining process tool calibration, determining wear status and check that solar modules are processed correctly.

The JENOPTIK-VOTANTM Solas 100/200 was designed for pattern scribing on thin-film solar modules using laser and/or needles for all the process steps – P1, P2 and P3. In the standard machines, six industrial cameras combine with VisionPro vision software from Cognex to maintain process-tool calibration, determine wear status and check that solar modules are processed correctly.

Because set-up, processing and the quality inspection takes just 60 seconds per module, manufacturers can achieve a throughput advantage in the highly competitive sustainable energy market.

Thin-film technology that relies on vapour-deposited or sputtered photoactive semiconductors on a glass substrate, generally use less energy and material, and offer reduced manufacturing costs when compared with silicon-based solar cells. Lower production cost allows manufacturers to make more comfortable pricing decisions.

Modern thin-film solar cells consist of a metal layer, a semiconductor layer and a transparent, electrically-conductive oxide layer. In the first step, the VOTAN Solas 100 uses several lasers to scribe patterns in the bottom layer, also called P1. On CIS/CIGS modules, mechanical tools process the two layers above this, P2 and P3. An integrated needle-comb unit precisely scribes the structures into the surface at a speed of up to 1.5m/sec. Each needle is individually actuated and positioned in this process, allowing the system to adjust quickly to different cell sizes and react accurately to changes. Automatic needle alignment minimizes setup times to achieve greater throughput.

The vision system records the actual position of the tools in seconds. At the same time, three additional cameras measure the P1 track of the automatically-fed solar panel. The cameras supply three reference points for auto-alignment and depending on the location and position of the P1 structures VisionPro provides data to adjust the machine’s coordinate system and correct the tool path. The goal of the process is to minimize displacement of the P1 structures in relation to the P2 and P3 layers to achieve cell efficiency. Processing tools then scribe the P2 and P3 layers to a positional accuracy of 5µm.

In the final inspection step, two more cameras closely examine the solar module again and inspect the quality with VisionPro vision software looking for micro cracks, broken glass and potential detachment of layer particles.

This machine, with VisionPro vision software, is said to be quick and easy to integrate into complex production lines. The software ignores non-critical changes in the appearance of the solar panel and concentrates on the features important for the quality of products. Also, the tools do not require complex image pre-processing, which accelerates application development and reduces lifecycle costs.

Control Engineering Europe