[PV Robot Front Line] Panel Inspection Robot by Atox (page 2)

Inspection services based on 'robot management' expertise

2014/06/20 13:05
Kenji Kaneko, Nikkei BP CleanTech Institute
Print Page

Images are recognized by a CCD camera for running, and, according to the explanation, the robot can move over the gap between the panels on its own if the gap is less than 10cm. It can run on panels angled up to 30° if the surface condition (cleanliness) is good. According to the explanation, the friction force of the incorporated rubber was measured, and the contact area of the crawler that prevents the 10kg main unit from slipping was calculated. However, the friction force declines in rainy weather, making it impossible to run.

The most characteristic part of the running method is the turning technique that lifts up the main unit. In vehicles equipped with crawlers such as tanks and heavy machinery, directions are normally changed by an "ultra-pivotal brake turn," where crawlers on both sides are turned in directions opposite to each other. This turning method, however, features high load on the road surface. If the ultra-pivotal brake turn method is incorporated in the robot that travels on solar panels, the glass surface could be scratched.

Because of this reason, a turning mechanism that lifts up and turns the entire unit was incorporated in the center of the main body (Fig 4). A disk that is attached by rubber and about 25cm in diameter is pressed against the glass surface, and the main unit is lifted up and turned 90°, preventing the glass surface from being scratched by the crawlers (Fig 5).

"This kind of turning method is frequently used in models introduced at robot contests," Yamada said.

Travels along vertical, horizontal lines of panel

The robot can travel on its own on the panel surface because the model number and cell layout information of the panel to be inspected are stored in advance as an "address" and the current position in the address is obtained by processing the images of the CCD camera. The camera recognizes lattice patterns on the panel surface and judges the vertical lines and the horizontal lines by image processing.

(Continue to the next page)