ISIA – Data Collection: Sensors

What are sensors?

  • Device, module, or system whose purpose is to
    • detect events or changes in its environment
    • send the information to another (electronic) system
  • Some examples:
    • touch-sensitive buttons (tactile sensors)
    • temperature, pressure, flow measurement
    • AHRS (attitude and heading reference system) for aircraft
      • gyroscope: roll, pitch yaw; compass: heading
      • ‘nine degrees of freedom’ input for Wii and game controllers
    • MARGS (magnetic, angular rate, and gravity sensor)
    • position (potentiometers), pressure (force/stress sensing resistors)
  • applications
    • manufacturing/machinery, aerospace, transportation, medicine, robotics

Miniaturization of Sensors

  • Micro-electro-mechanical systems (MEMS)
    • in Europe: ‘micro systems technology’ (MST)• in Japan: ‘micromachines’
  • Microscopic devices with moving parts
    • internal components between 1 and 100 micrometres in size (0.001 mm to 0.1 mm)
    • entire device typically between 20 micrometres (0.02 mm) and 1 mm per side

MEMS gyroscope

  • Microscopic ‘tuning forks’
  • Coriolis force lateral to plane of rotation) change in vibration mode

MEMS devices and applications

  • Accelerometers
    • airbag deployment, vehicle stability control
    • camera drop detector
  • Gyroscopes
    • roll, pitch and yaw in aircraft AHRS and autopilot
  • Magnetometer
    • directional heading, magnetic field detector
    • metal detector (sunken ships), mining hazards
  • Microphones
    • portable telephones, headsets, laptop computers
  • Pressure sensors
    • car tire monitoring, blood pressure measurement, barometers

Sensor characteristics

  • Sensitivity
    • output change relative to measured input quantity change
    • law: linear, log, exponential
  • Range
    • input range: min. and max. true values measurable
      output range: min. and max. output values that can be generated
  • Resolution (discrimination)
    • the minimal input change needed to cause a detectable output change

Accuracy and precision

  • Accuracy
    • how close the output is to the true value being measured
      absolute error = output true value
      relative error = absolute error ÷ true value
  • Precision
    • related to the variance in a set of measurements of the same true value
      • measurements not necessarily accurate
    • repeatability = precision of a set of measurements over a short time interval
    • reproducibility = long-term precision, or between operators/sensors/environments

Sources of error

  • Systematic errors (bias)
    • temperature instability
    • drift (temperature, mechanical, chemical)
    • loading errors (the sensor affects the true value)
    • attenuation (lossy transmission channel)
    • human observational errors (e.g., parallax)
  • Random errors (noise)
    • natural fluctuations in the true value (e.g., water height, with waves)
    • background noise (e.g., fluorescent lights causing ‘mains hum’))
      → signal-to-noise ratio (SNR) should be >> 1

Static input/output relationships

  • Linearity
    • closeness of the output to an ideal line or curve
      ? easily compensated for
  • Monotonicity
    • increase in true value → increase in measurement
      ? non-monotonic output → ambiguous measurements
  • Hysteresis
    • dependence of the output on the input history
      ? loose mechanical coupling (‘backlash’) → ambiguity
      ? electrical hysteresis improves digital noise immunity


  • Transducer
    • device converting signal from one physical form to another physical form
  • Sensor = input transducer
    • receives and responds to a signal or stimulus
    • typically converts physical input signal to electrical output
    • e.g., microphone
  • Actuator = output transducer
    • generates a signal or a stimulus
    • typically converts electrical input signal to physical output
    • e.g., loudspeaker


  • Comparison of sensor measurement values with those of a standard of known accuracy
    • another measurement device of known accuracy
    • a device accurately generating the quantity to be measured
    • a physical artifact, such as a standard meter ruler
  • Usually followed by
    • certification of the device under test
    • the measured error being noted
    • adjustment made to correct the error to an acceptable level

Automatic range calibration

  • Continually re-calibrate and adjust input range
    • remember maximum and minimum input values (the input range)
    • update input range if new input value is outside previous range
    • adjust output scaling appropriately for current input range

Useful when output range is fixed regardless of input range

Simple instrument systems

  • Processor typically preceded by analogue-to-digital converter (ADC)
    • electrical signal (voltage) converted to digital value (integer)
    • fixed range (input voltage) and resolution (number of output bits)
  • ADC can be a source of errors
    • non-linearity (missing output values)
    • non-monotonicity (repeated output values)


  • Often integrated with input ADCs
    • calibration, scaling, conversion, etc.
    • low-power and small size desirable
    • communication and/or local display
  • Integrated sensor nodes combine
    • input sensor
    • ADC and processor
    • communications (Wi-Fi, ethernet, USB)
    • protocol stack (TCP/IP, HTTP, etc.)

Practical sensing

  • Build a simple instrument system
    • sensor (e.g., light, temperature) → voltage
    • processor (embedded CPU) voltage → ADC → digital value
    • display (computer screen or numeric display)

Environmental sensing: measuring light level

A photocell is a resistor whose resistance depends on light level

  • Light-dependent resistor
    • more light → less resistance
  • How do we turn a changing resistance into a changing voltage?
    • hint: compare with potentiometer (variable resistor)