Top to Bottom: Laser scan of a tilted surface, video measurement of the edge of a paint spot and touch probing of the cylinder wall are all possible on a properly configured multisensor measurement system with a single part set up. Source: Optical Gaging Products Inc.

Precisely measuring dimensions of manufactured parts is fundamental in manufacturing quality control. Increasingly, companies are relying on multisensor measurement systems for measurement verification.

But how can a machine with several sensors be more efficient? It helps to understand what each sensor can do.

  • Video Measurement.
  • Video measurement is simple in concept-image the part with optics onto a camera then measure the image. The image must be in focus to measure accurately. Autofocus can focus points on a surface or edges of features. A zoom lens can magnify the image so details too small to probe with a coordinate measuring machine (CMM) can be collected, such as points along a critical radius or the perimeter of a slot. And, unlike single point touch trigger probing, numerous points can be collected simultaneously.

    Video measurement of manufactured parts is not limited to what can be seen in one camera image. By accurately moving the part and/or the optics and tracking the position with high-resolution scales, the advantages of video measurement can apply to parts with dimensions a meter or more apart.

    Some people assume video measurement is limited to flat parts or measurements in a single plane. This is no longer the case. The ability to focus anywhere in the machine volume allows the software to present the dimension and angle relationships between those points. Video measurement is fast and noncontact, only limited to measuring what the camera can image.

  • Touch Probes.
  • Touch trigger probes are another staple of manufacturing quality control. Touch trigger probes are most associated with CMMs but are a common sensor addition for multisensor video measuring systems.

    Touch trigger probes trigger with part contact. Since a probe can trigger from surface contact from any direction, it can measure points on features that video cannot image, such as inside a cylinder bore.

    A disadvantage of touch trigger probes is that measuring numerous points is slow. Each point is sequentially touched and recorded so it takes time to touch enough points to characterize a feature in detail. Fragile or deformable parts may also cause a problem due to part deflection.

  • Laser Measurement.
  • A laser sensor adds another form of noncontact measurement. When added to a video-based multisensor system, the laser can be mounted on-axis with the video optical system (TTL–through-the-lens), or on a separate axis, parallel to the video optical system. The laser light is accurately focused on a surface yielding a point location just as a touch probe does.

    The advantages of a laser sensor are that the light can be focused to a very small area, the technique is noncontact and data point acquisition rates are very high. This makes this sensor ideal for measurements such as flatness and Z measurements on fragile parts.

    Coordinating Sensors and Measurement

    The multisensor system’s metrology software controls all the sensors, tracks their positions within the system’s measurement volume, including any offsets from one another, and controls data acquisition.

    For true 3-D measurement the software maintains the positions of all acquired data points within the measurement volume, allowing those points to be used for constructing relationships that are measurable.

    Using Sensors Correctly

    To get the best results from a multisensor measurement system, it is important to use the best available sensor for each part feature. For example, video excels at measuring edges. An edge can be the perimeter of a part, the top of a bore or a slot or a group of features such as a bolthole pattern. In fact, an edge can be the boundary between areas of different colors or textures on a surface.

    Metrology software easily measures features that fit in one camera field of view, while the more capable metrology software applications can automatically move the camera or part to follow an edge much larger than the camera image, measure multiple features simultaneously, or measure the absence of material noting the area, dimensions and center of the “blob” of pixels those data represent.

    Although video is capable of measuring surface points, a laser may be faster and more accurate. Besides collecting a surface point at a time, some lasers can be scanned over surfaces, acquiring points as they move. When integrated into the system’s overall stage motion control, the laser can scan greatly varying, complex curved surfaces.

    A touch probe can add data points from part surfaces inaccessible to laser or video while the part is in a fixed position. For example, the probe can collect points from the walls of a bore that is on the same axis as the optical system. The optics can easily measure the top circle where the cylinder intersects with the top plane but it cannot measure the walls of the cylinder. The probe can acquire points around the cylinder at several depths from the top to measure any taper or asymmetry of the cylinder.

    Video, laser and touch probe are the major types of sensors used in multisensor metrology. There are micro-probes that use unique triggering technologies. And some touch probes can be scanned, similar to scanning a laser, not only able to scan a top surface, but also perpendicular surfaces.

    The benefit

    Rather than accept incomplete information, or delay important manufacturing decisions due to measurement system limitations, a computer numerical control (CNC) multisensor system can switch sensors during an automatic routine providing thorough, detailed measurements with little or no operator interaction other than loading and unloading parts. Even the most complex multisensor measurement routine may be faster than manual measurements on a single-sensor system, boosting the efficiency of the measurement verification process.