Lightway Systems: Customs Designed and Built Optical-Based Industrial Inspection Sytems

Custom Optical
Sensors and Inspection Systems

How can I measure surface finish (roughness)?

How can I measure surface color?

How can I detect cracks, flaws, and imperfections?

How can I do dimensional checks?

How can I detect vibration?

How can I monitor concentration and composition of liquids or gels?

Can a single system handle multiple sensors?

How will a Lightway's system integrate into my existing control protocol?

Can I see a summary of optical inspection applications?

What are the main roughness parameters?

What are the "land mines" in the optical dimensional measurements?

How can I measure surface finish (roughness)?

Optically, surface roughness can be measured in both relative and absolute units of measurement, but only in its statistical form - Ra.

Relative measurements produce results proportional to the variance of the surface finish of the actual process components. Translation from relative to absolute units is possible via calibration of the sensor's output against a "reference standard". These measurements are convenient in a moving-process, unmanned checks because of its moderate cost and relative immunity to the position of the surface under measurement.

Measuring roughness in absolute units such as microns or microinches is also possible, but requires a precision sensor with an associated cost premium. Such sensors are used in industries where the complexity and cost of the measurement equipment is compensated for by the results achieved, e.g. semiconductor or hard disk drive manufacturing.

The optical methods has several key advantages: (a) areas can be measured in situations where mechanical styluses cannot access, (b) sensitivity of the method is much higher than that of stylus tips, (c) non-contact, hence, non-destructive and clean.

How can I measure surface color?

Top of Page

Surface color measurement is the area of photometry where everything is related to the response of human eye to the spectral distribution of light. Numerically, a color can be described in a "color space" - a set of functions with 3 parameters. To accommodate various existing industrial equipment working with colors, different types of color spaces have been developed - RGB, Lab, CMYK, HSB, XYZ, etc. Any color can be translated between various color spaces.

The sensor / light source arrangement is critical to performance; Lightway Systems will provide the expertise to ensure a successful application.

How can I detect cracks, flaws, and imperfections?

Top of Page

This is an ideal application for using a polarized light. Even tiny micron-size surface imperfections will distort the expected polarization state of the reflected light. A sensor camera will capture the distortions and show them vividly.

How can I do dimensional checks?

Top of Page

Machine vision is the convenient answer to the problem. Typically for industry, the practical (and sometimes theoretical) limit of sizes handled by vision-based systems is 2-3 microns. System components must be selected according to the required performance parameters such as; processing speed, surface area, image size, etc. Once again, the expertise of application will ensure the right components for the right task.

Measuring sizes comparable or below the wavelength of light can be conducted reliable with interference methods. This raises the bar of standards for material handling, optical precision, and the final cost.

How can I detect vibration?

Top of Page

A slightest vibration can be easily detected utilizing an optical fiber / sensor arrangement. When stressed, a twisted optical fiber modulates the intensity of light passing through it; a sensor is positioned to detect and measure the degree of vibration.

How can I monitor concentration and composition of liquids or gels?

Top of Page

Using a sensor designed to measure the refractive index either on-line or via samples both concentration and/or composition can be monitored.

Can a single system handle multiple sensors?

Top of Page

A single industrial computer with Lightway's sensor interface board(s) can simultaneously handle the output from multiple sensors, for either separate process control or consensus decision making. Multiple sensors can be used for redundancy or logic applications.

How will a Lightway system integrate into my existing control protocol?

Top of Page

Protocol compatibility is achieved through the use of commercially available interface cards. Thereby ensuring a seamless integration into any existing architecture. Can I see a summary of optical inspection applications?

Can I see a summary of optical inspection applications?

Top of Page

CATEGORY	PROBLEMS TO SOLVE
Image Analysis	Part Recognition / Sorting Presence / Absence Surface pattern statistics Dimensional Measurements Texture analysis
Surface Properties Analysis	Color characteristics Finishing (roughness) "Skin layer" uniformity Presence / Condition of coatings Surface profile statistics
Material Properties Analysis	Refractive index Optical constants Stress Mixture composition / uniformity Thickness

PROPERTY OF LIGHT	WHAT TO INSPECT BASED ON ...
PROPERTY OF LIGHT	TRANSMISSION	REFLECTION	AMBIENT RECEPTION
Intensity	Optical Density Absorption Thickness	Distance Roughness Temperature	Intensity Pollution Temperature
Spectral Characteristics	Color Pollution Material Properties	Color Temperature Humidity Material Properties	Light Sources Pollution Flame Detection Temperature
Polarization State	Stress	Material Properties Thin Film Thickness	Pollution
Spatial Distribution (image, specularities)	Pattern Dimensions Texture Flaws	Pattern Dimensions Texture Surface Profile Flaws Roughness	Dimensions Distance
Coherence	Thickness Distance	Distance

What are the main roughness parameters?

Top of Page

1. Rq - Root Mean Square Roughness
2. Ra - Average Roughness
3. Rz - Average Maximum Height of the Profile
4. Rp - Maximum Profile Peak Height
5. Rv - Maximum Profile Valley Depth
6. S - Mean Spacing of Local Peaks of the Profile
7. Sm - Mean Spacing of Profile Irregularities

What are the "land mines" in the optical dimensional measurements?

Top of Page

The main limitations are:

1. Range-to-resolution ratio: Any array of photo detectors, linear or 2D, has finite number of pixels. Therefore, (a) the maximum achievable resolution is RANGE / NUMBER-OF-PIXELS, or (b) the maximum range is REQUIRED-RESOULUTION * NUMBER-OF-PIXELS. Attempts to defeat this trade-off with more complex analysis software techniques like "pixel interpolation" are somewhat successful and must be used with care due to the fact that mathematical procedures, while improving on resolution, may cause undesired side effects like shape distortion, Moiré noise, etc.

2. Optical aberrations: if the lens used does not "picture" the image precisely, it is impossible to conduct quality measurements no matter how good the camera and image analysis software are. This explains the price gap between the consumer lenses and the gauging lenses.

3. Lighting: it is critical and not any less important than quality of optics and electronics. Lighting must correspond to the type of optic used whether it is telecentric, bright field, dark field, etc. Without proper illumination, in the case of defect or pattern recognition, the camera becomes less sensitive or even blind to what it supposed to watch for. In dimensional measurements, unevenness in illumination can cause dimensional distortions.

4. Timing: it takes time to (a) take a frame, (b) bring the image into a software-accessible memory, and (c) analyze the image. Most of the current frame grabbers combine (a) and (b). The use of high-speed cameras reduces the time to acquire an image, but there is a trade-off of the low resolution (typically, 640x480 pixels). In the case when multiple cameras are used, the situation gets more complex - often cameras must be synchronized in order not to have "gaps" in inspection processes. Image analysis can be sped up by optimizing software routines and reducing their amount to only what is absolutely necessary.