Metrology: The Science of Measuring Physical Quantities

Geometric quantities describe the spatial properties of a physical object. 4 geometric quantity types dominate industrial measurement: form (surface shape and deviation from nominal geometry), surface texture (roughness, waviness, and primary profile), position (location and orientation in a coordinate system), and dimension (length, width, diameter, and angular extent). Industrial 3D sensors measure geometric quantities by projecting structured light or laser lines onto a surface and recording the reflected pattern with a CMOS or CCD image sensor.

Surface roughness Ra — the arithmetic mean deviation of the surface profile from the mean line — is one of the most frequently specified geometric quantities in technical drawings and quality standards. The relationship between profile height $z(x)$ and the Ra value is defined as:

Ra=1l∫0l∣z(x)∣dxRa​=l1​∫0l​∣z(x)∣dx

where $l$ is the evaluation length and $z(x)$ is the profile height at position $x$.

Thermal quantities describe the temperature state and heat transfer behaviour of an object. Temperature is the most frequently measured thermal quantity in industrial inspection; thermal emission (infrared radiation intensity as a function of surface temperature) is the second. Infrared cameras measure thermal quantities by detecting long-wave infrared radiation in the 8–14 µm spectral range without contact with the measurement object. Thermal measurement detects component overheating, weld seam defects, and thermal bridges in building envelopes.

The radiated power per unit area of a surface follows the Stefan–Boltzmann law:

M=ε⋅σ⋅T4M=ε⋅σ⋅T4

where $\epsilon$ is the surface emissivity (dimensionless, range 0–1), $\sigma$ is the Stefan–Boltzmann constant 5.67×10−8W m−2K−45.67×10−8W m−2K−4, and $T$ is the absolute surface temperature in kelvin.

Optical quantities describe the interaction of electromagnetic radiation with surfaces and materials. 3 optical quantity types appear in industrial inspection: luminous intensity (optical power per unit solid angle), spectral reflectance (wavelength-dependent surface reflection coefficient), and transmittance (fraction of incident radiation passing through a material). Multispectral and hyperspectral sensor systems measure spectral reflectance across defined wavelength bands to classify material composition and surface condition.

Mechanical quantities include force (measured in newtons), pressure (measured in pascals), and strain (dimensionless deformation ratio). Time and frequency quantities describe process duration, cycle rates, and vibration frequency. Industrial measurement systems additionally address electrical, acoustic, chemical, and biological quantities; these quantity categories are treated in their respective dedicated articles within this measurement knowledge base.

Laser triangulation is an optical measurement principle that determines surface distance by measuring the angular displacement of a reflected laser point or line relative to a known baseline between the laser emitter and the image sensor. A 3D profile sensor projects a laser line onto the measurement object; the image sensor captures the reflected line profile at a defined triangulation angle. Distance resolution in laser triangulation sensors reaches values below 1 µm for short measurement ranges.

The triangulation distance $d$ is calculated from the triangulation angle $\alpha$, the baseline length $b$, and the displacement $s$ of the laser spot on the detector:

d=b⋅fsd=sb⋅f​

where $b$ is the baseline between laser emitter and detector, $f$ is the focal length of the receiver optics, and $s$ is the spot displacement on the detector array. If dˉ(x,y)dˉ(x,y) increases monotonically with increasing distance from the sensor, the measurement is unambiguous within the defined measurement range.

Structured light projection extends laser triangulation to full-surface 3D measurement by projecting a sequence of fringe patterns and computing a depth map from phase shifts in the deformed pattern.

Infrared thermography is a thermal measurement principle that converts the infrared emission of a surface into a calibrated temperature map. Every object above 0 kelvin emits infrared radiation; the emitted power per unit area follows the Stefan–Boltzmann law as a function of surface temperature and emissivity. An infrared camera captures the spatial distribution of this emission with a focal plane array (FPA) detector — either an uncooled microbolometer or a cooled photon detector — and converts detector response into temperature values using a factory-calibrated conversion function.

Thermographic measurement detects thermal gradients, subsurface defects, and heat loss patterns without mechanical contact with the component.

Reference standard comparison is a measurement principle that determines the value of a measurand by direct comparison against a physical reference object with a known, traceable value. Gauge blocks, ring gauges, and calibration standards serve as reference artefacts. Geometrical-optical methods — including confocal microscopy and interferometry — measure surface topography by analysing the optical path length difference between a measurement beam and a reference beam. These methods achieve sub-nanometre depth resolution on specular surfaces.

Electrical measurement principles — resistive, capacitive, inductive, and piezoelectric transduction — convert mechanical or thermal quantities into electrical signals. Capacitive sensors measure displacement and film thickness by detecting changes in electrical capacitance between two conductive surfaces. Inductive sensors detect the proximity of ferromagnetic objects through changes in magnetic flux linkage. Magnetic, acoustic, mechanically analytical, and chemically analytical measurement principles constitute further principle categories applicable to specialised industrial inspection tasks.

The measurement object is the physical item on which the measurement is performed. Measurement object properties — including surface reflectance, thermal emissivity, material homogeneity, and geometric complexity — determine which measurement principle and instrument are applicable. Contamination, vibration, and thermal expansion of the measurement object are 3 primary environmental influences that introduce systematic errors into the measurement result.

The measurand is the specific physical quantity subject to measurement. Correct measurand definition requires specifying the quantity type (e.g. surface roughness Ra in micrometres), the measurement location on the object, and the applicable measurement standard (e.g. ISO 4287 for surface texture parameters). Ambiguous measurand definition is a systematic error source that invalidates comparison of results across measurement systems.

The measurement instrument is the device that detects the measurand and converts it into a readable signal or data output. 3D sensors, infrared cameras, coordinate measuring machines (CMM), and laser displacement sensors are 4 instrument types used in industrial dimensional and thermal measurement. Instrument selection criteria include measurement range, spatial resolution, acquisition speed, and environmental robustness rating (IP protection class).

The measurement procedure specifies the sequence of operations, instrument settings, reference positions, and evaluation algorithms applied to obtain the measurement result from the raw sensor signal. Standardised measurement procedures — defined in ISO, VDA, or manufacturer-specific guidelines — ensure reproducibility across operators and measurement cycles. Laser triangulation profile measurement procedures define scan speed, laser power, ambient light suppression, and peak detection algorithm.

Measurement conditions are the environmental and operational parameters that influence the measurement result independently of the measurement object and instrument. 5 measurement condition categories affect optical and thermal measurement systems: ambient temperature (thermal expansion of instrument and object), ambient light (stray light interference in image sensors), vibration (image blur and displacement error), humidity (optical surface condensation and thermal emissivity shift), and electromagnetic interference (signal noise in detector electronics).

The measurement result is the output value obtained after applying the measurement procedure to the raw sensor data. A complete measurement result includes the measured value, the associated measurement uncertainty, the reference standard used, and the measurement conditions at the time of acquisition. Measurement results serve as the data basis for 4 evaluative outputs: classification decisions (pass/fail), process alarms, statistical process control (SPC) values, and quality documentation records.

The combined standard measurement uncertainty \( u_c \) is calculated from the individual standard uncertainties \( u_i \) of all input quantities \( x_i \):

\[ u_c(y) = \sqrt{\sum_{i=1}^{N} \left(\frac{\partial f}{\partial x_i}\right)^2 u^2(x_i)} \]

where \( \frac{\partial f}{\partial x_i} \) is the sensitivity coefficient of the output quantity \( y \) with respect to input quantity \( x_i \), as defined in the GUM (Guide to the Expression of Uncertainty in Measurement).

Optical sensor systems capture geometric and surface information by analysing the interaction of structured electromagnetic radiation with the measurement object. 3D sensors based on laser triangulation measure surface profiles at acquisition rates exceeding 10,000 profiles per second — a performance parameter that enables inline 3D inspection at production line speeds. 3D imaging systems extend profile measurement to full-surface point cloud acquisition, generating dense 3D data sets with lateral resolutions below 50 µm and depth resolutions below 10 µm on industrial components.

3D sensors acquire the three-dimensional spatial coordinates of surface points by computing depth from structured light deformation, stereo triangulation, or time-of-flight measurement. 3D imaging systems integrate 3D sensor hardware with embedded processing and software evaluation to output measurement results — dimensional values, surface maps, or pass/fail decisions — directly from the sensor. On-sensor processing reduces data transmission volume and latency compared to PC-based evaluation architectures.

Infrared sensor systems measure thermal emission in the spectral range from 1 µm to 14 µm to determine surface temperature distributions and detect thermal anomalies. 2 infrared detector technologies dominate industrial thermographic measurement: uncooled microbolometer arrays (operating at ambient temperature, spectral range 8–14 µm) and cooled photon detectors — InSb or MCT (mercury cadmium telluride) — operating in the mid-wave infrared range (3–5 µm) at higher sensitivity and frame rates.

Image processing sensors with on-sensor processing integrate image acquisition, feature extraction, and measurement evaluation into a single sensor unit. On-sensor processing applies preprocessing algorithms — including intensity normalisation, peak detection, and sub-pixel interpolation — directly on the sensor’s embedded processor before transmitting structured measurement data to the host system. This architecture reduces raw data volume by a factor of 10 to 100 compared to unprocessed image transmission and supports cycle time requirements below 1 ms in high-speed inspection applications.

Electrical and magnetic sensor systems — Hall-effect sensors, inductive proximity sensors, and capacitive sensors — measure position, displacement, and field strength in applications where optical access to the measurement object is restricted. Mechanical sensor systems, including strain gauge arrays (DMS) and MEMS-based accelerometers, measure force, pressure, and vibration. Acoustic, chemical, and biological sensor systems address measurement tasks in process analytics, environmental monitoring, and life science instrumentation; these sensor system categories are treated in their dedicated articles within this measurement knowledge base.