Honda NSX robotic inspection system

A tier-two automotive client of ours had a requirement to automate their inspection process for 15 variants of aluminium castings for the Honda NSX supercar range. These castings are critical to the construction, strength and rigidity of the vehicle. The solution supplied was a fully automated high accuracy robotic inspection system using technologies such as vision, 3D scanning and micrometre probe testing. The solution was housed within a darkened cell to provide a repeatable and consistent background to ensure we achieved the most accurate images. The cell was made on a reinforced frame to ensure a repeatable low vibration environment for consistent inspection results.

A Fanuc –M10iA/12S 6 axis industrial robot with a repeatability of less than 40µ, was used to manipulate a casting in all orientations to present to a series of Keyence laser displacement sensors and measurement probes. Two high-resolution OP-To and Allied Vision 30-megapixel telecentric lenses and Dalsa cameras capable of 10µ of accuracy to inspect; machined surfaces, presence of rivnuts, rivets, pop nuts, HB bolts and stud rivets. The system will compare the current casting inspection results to a referenced “Golden part” stored within the pre-programmed database. This will determine whether the casting will pass or fail.

The overall system accuracy was achieved at less than 70µ for each variant. The variant sizes varied significantly with the largest part greater than 400mm x 350mm and the smallest part 150mm x 175mm. To achieve an overall accuracy of <70µ and considering that the robot equated to 40µ of the total 70µ. QM Systems required a 30-megapixel camera with a huge telecentric lens in order to gain a total accuracy of under 70µ. One of the drawbacks of this system was that, to gain this accuracy we could only do so with a field of view of 100mm². This meant that a complex algorithm was required in order to re-datum each robot movement and create new datums referenced from the existing known datum point. We had to consider a stack-up of inaccuracies and their tolerances to be incorporated in our calculations. This project was a blue-sky thinking project that pushed the boundaries of technologies. QM systems developed a working system defying many industry specialists who said it couldn’t be done.
QM Systems Innovation in Automation.