Humidity Sensor

A humidity sensor measures and reports moisture and, optionally, air temperature.

The typical observation variable in this context is relative humidity, which is the ratio between the current moisture in the air and the highest amount of moisture at a particular air temperature, which is a relevant variable both in industrial processes as well as in personal comfort. Depending on the employed mechanism, there are three classes of humidity sensors: capacitive, resistive, and thermal. To begin with, the capacitive sensors are based on oxidation; they read the level of oxidation of a strip of metal placed betw-een two electrodes. Given the employed mechanism is a nonlinear chemical one, this category requires significant processing and normalization to ensure a linear reading. The resistive sensors rely on the presence of ions in a salt medium present betw-een two electrodes; changes in humidity will also the concentration of ions and, subsequently, the resistance of the electrodes. Finally, the thermal sensors observe the variation in resistance between two electrodes, one in dry nitrogen and one at ambient temperature.

The humidity sensors have a number of defining parameters, from the accuracy and linearity to reliability. In addition, their drift is measured through repeatability and, given they are based on a relatively slow chemical process, the response time once the stimulus is applied.

While the importance of humidity is less obvious from a human perspective, apart from the level of comfort, measuring it is critical for industrial and agricultural applications. A few particular areas to highlight are the textile industry, as humidity changes the fabric properties, and agriculture, to determine the level of moisture in the air and predict the quality of these. [40].

Accelerometer

An accelerometer is a type of sensor that detects moving acceleration in terms of amplitude and frequency. Traditionally accelerometers have been used in relation to monitoring human movement, from observing possible movement issues (from motor recovery to fall detection) to detecting and improving posture and gait. With the advent of wearable devices, accelerometers have been used in a range of monitoring and classification of activities, from sitting or standing to running or cycling.

Accelerometers are conceptually based on a damp mass on a spring. Given an external acceleration on the direction of the spring, the mass will expand or contract the spring proportionally with the strength of the acceleration. Depending on the type of response, there are two types of accelerometers—DC response and AC response. The first category measures static accelerations, such as gravity, that do not vary in time, hence have zero variation. АС-response accelerometers can observe variation in acceleration (such as the one produced by a hand movement) or vibration.

Implementation-wise, there are three types of accelerometers: capacitive Micro- Electro-Mechanical Systems (MEMS) [42], piezoresistive, or piezoelectric. As then- name indicates, MEMS are a combination of electrical and mechanical component which transforms the capacitance of a mass that is subject to acceleration. Due to their small factor, MEMS are very suitable for integration into surface-mount devices or printed circuit boards; in terms of the above classification, they are DC-coupled. The miniature size does impact on their accuracy, as measurements may be rather noisy and allow a rather narrow measurement range. Piezoresistive accelerometers are also DC-coupled, but rely on the changes in resistance when acceleration is applied to the sensor mass. They benefit from a wide measurement range, so can measure high-impact events, such as shocks, but have a low sensitivity, hence should not be relied upon for accurate measurements. Because of their construction characteristics, piezoresistive sensors are the most expensive category of accelerometer. Finally, piezoelectric sensors rely on the property of lead zirconium titanate to produce an electrical charge when acceleration is applied to it. This category of sensors benefits from high sensitivity, low noise levels, but the range of applications is slightly limited due to their AC-coupled nature and can saturate if subjected to an acceleration outside their measurement interval. The accelerometer sensor suitability for different applications is presented in Table 2.2.

Proximity Sensor

One of the requirements of IoT is to allow more natural, human-oriented interaction. One option to achieve this is to augment sensing from switch-, action-based to observational. A basic example is sensing the presence of a user rather than requiring them to interact with a system, undertake a task, or press a switch. Proximity sensors are one of the distinctive categories fulfilling this requirement, as they convert the presence of a subject into an electrical signal. From a physics perspective, there are four underlying mechanisms: inductive, electrical capacitive, magnetic, and ultrasonic. To start, the first category monitors the eddy currents generated in a metallic sensing object due to electromagnetic induction; second type of sensors observe variations in

TABLE 2.2

Summary of LPWAN Technologies

electrical capacity when approaching the sensing object, the third category, as per its name, use magnets and reed switches, and the last one observes the reflecting sonic waves as they bounce off a detected object. Due to the respective physical phenomenon observed, the three categories also react to different stimuli. The inductive sensors will detect the presence of a metallic object, the capacitive sensors will detect the presence of a wider range of materials starting with metallic objects down to resins or powders, the magnetic sensors will only detect magnets, while the ultrasonic sensors can be used for distance measurement and monitoring. Given their sensitivity to variation in electrical charge or current, the first two are very susceptible to electrical noise, from power lines in their vicinity down to long cables between the sensing element and the interpreter within the sensor.