Opto presents Automated Ball Bearing Analysis Using Light Reflections

In industry, precision is considered a sacred virtue, especially when it comes to ball bearings. But while the master’s eye is known to inspect well, it is not sufficient to detect micro-scratches and variations in surface roughness. This is where our Clusterpartner Opto’s Solino algorithmic technology comes into play: a system that uses light reflections to relentlessly uncover even the tiniest defects.

Optical Measurement Technology

Image 1 | Solino is an optical sensor, among other things, for ball bearing inspection. With algorithms adapted to specific bearing defects, the desired classifications can be consistently reproduced and evaluated using AI. – Image: Opto GmbH

Ball bearings require an extremely smooth surface to ensure low-friction movement. Increased roughness leads to uneven operation, accelerating wear and material degradation. Irregularities or the smallest scratches add extra friction to the surface. Additionally, ball bearings are often coated with a thin protective layer or lubricants. Coating defects or contaminants such as tiny dust particles can prevent lubricants from adhering or distributing properly. Even microscopic cracks can develop in the contact areas.

Measuring Light Scattering

The Solino imaging module from Opto is an optical inspection sensor particularly suited for highly polished surfaces like ball bearings. It analyzes how light is reflected from the surface and can detect micro-scratches, roughness variations, coating defects, and contaminants with high precision. Multiple LEDs project light onto the bearing at specific angles, while a 20MP camera captures the reflected light from different angles. With field sizes of 10×10, 25×25, or 50x50 mm, various sensors are available, allowing the system to be tailored to the application’s resolution and defect detection needs. A flawless surface reflects light in a predictable pattern, whereas defects scatter light irregularly. The Solino algorithm compares the measured reflection profile to an ideal, defect-free reference. To capture every possible defect, the LEDs are arranged to ensure no scattered reflection is missed. Even light striking at shallow angles is considered, resulting in a short working distance of just 5 mm. Deviations in the reflection pattern indicate defects, such as:

• Coating defects (changed reflection intensity)

• Roughness variations (diffuse reflection)

• Contaminants (irregular light scattering)

• Micro-scratches (uneven scattering)

Image 2 | Solino is a plug-and-play AI vision sensor in a rugged design. – Image: Opto GmbH

Surface Irregularities Below 1 µm

Even the tiniest surface irregularities (<1 µm) measurably alter the reflection pattern. Solino detects defects invisible to the human eye or standard cameras. It can even identify flaws smaller than resolution limits, as their scattered light impulse can still be evaluated as a signal. The algorithm also highlights changes in surface reflection that may indicate stress fractures before they become critical. Traditional visual methods struggle to differentiate between normal polishing marks and actual defects. Solino provides a quantitative reflection profile that enables a clear distinction between acceptable and defective bearings. Importantly, the bearings are not mechanically stressed or damaged during inspection. Compared to other common non-destructive testing methods like laser or white light interferometry, eddy current, magnetic particle inspection, acoustic emission testing, vibration analysis, or thermography, Solino offers distinct advantages—especially for ball bearings with ultra-smooth surfaces.

The data can be seamlessly integrated into machine learning models and processed using AI algorithms to automatically classify and predict defects. This enables rapid and objective quality control in mass production. While Solino itself is not a measuring instrument, it offers reliable, repeatable, and 100% dependable visualization at high speed, while being robust against stray light.

For more information visit our Clusterpartner website.