Sharing is caring!

The Internet of Things and edge computing has contributed to the introduction of smart technologies into many areas of life: smart cities, smart factories, smart agriculture, smart medicine, and more. The basis of these technologies is the collection of data in real-time using various sensors (learn more here) and the subsequent analysis of the data obtained.

The number of IoT connected devices is projected to reach 64 billion by 2025, and the IoT market will reach $ 949.42 billion in the same year. The ever-accelerating growth in the production of various types of sensors also speaks to the growth of the Internet of Things industry. The hardware and software upgrade of sensors and systems based on them is happening at a rapid pace, and it is expected that in the next 25 years the size of sensors will be significantly reduced, and the sensors themselves will become even smarter and cheaper, which in turn will increase their use.

IoT cloud servers and edge (gateway) devices depend on sensors to collect real-time data. Since the reality around us, as a rule, operates with signals presented in analog forms, such as temperature in degrees Celsius, distance in meters, speed in kilometers per hour, pressure in newtons per square meter, etc., the task of sensors is to capture changes in these parameters in the environment and then digitize the resulting data.

Sensors are the endpoint in the Internet of Things network, as they are located further from cloud servers than other devices on the network. Although sensors are small and not as important to the network as cloud servers, they can play a critical role in system design and operation. One example of the importance of the sensor in the system is the recent crashes of Boeing 737 Max aircraft, where one of the factors that led to both tragedies was precisely a sensor malfunction.

Sensors must connect and interact with edge devices and cloud servers to operate on the IoT network. Currently, wireless technologies such as Bluetooth, NFC, RF, Wi-Fi, LoRaWAN, and NB-IoT (cellular communication) are mainly used to organize the interaction with servers and edge devices. The organization of the network and the number of sensors in it determine the required number of edge devices that are responsible for performing edge calculations (analyzing data from sensors) before sending information to cloud servers.

There are many types of sensors in various shapes and sizes that can measure almost any physical quantity. Some of them are one single component, for example, a light-sensitive diode; others – a module with various peripherals and an embedded microcontroller.

The sensors can be used to measure light, sound, temperature, pressure, position, changes in altitude and distance, the gas composition of air, speed and direction of movement, density and composition of a liquid, and so on. There are also many technology-based sensors used to detect and recognize objects: radar, LiDAR, light detector, magnetic detector, infrared (IR) detector, inductance-based sensors, imaging devices, ultrasonic devices, sonars, devices for working with photon radiation, touch recognition, encoders and much more.

Below is an approximate list of currently available sensor types:

  • Atmospheric pressure sensor
  • Distance / Proximity sensor
  • Humidity sensor
  • IR sensor
  • Level sensor
  • Light sensor
  • Motion Sensor
  • Smoke and gas detector
  • temperature sensor
  • Touch sensor
  • Ultrasonic sensor

Among other things, sensors can also be classified as active and passive, analog, and digital.

Active sensors usually require external support to operate. This can be an external power supply or inductance-based wireless power transmission. An example is a linear displacement differential transformer (LVDT) that can be used to convert the linear displacement to equivalent electrical signals: when working with an LVDT, energy comes from linear displacement through inductors without an external power source. Passive sensors do not need stimulation to work, for example, a thermocouple can convert heat directly into electrical signals and without additional energy sources.

How IoT sensors work with information

Any sensors collect analog data. Such data is continuous – it can be represented as a winding line, a continuous flow of information. You cannot transmit such data via cable or wireless communication – first, the signal must be converted into digital data.

Analog data can also be transmitted, for example by radio. But computers only work with digital data, so they still have to be converted to digital. And it is better to do this before transmission in order to use more modern and faster communication channels.

Simple analog sensors cannot convert a signal. To get information from them in digital form, a scale is needed where analog values ​​correspond to digital ones.

Example: Let’s take an ordinary mercury thermometer. As the temperature rises, the mercury expands. Mercury expansion and contraction information – analog data. To understand how many degrees outside, you need to impose a human-readable scale on the mercury column. And in order to transmit this data to other devices, connect a converter with a built-in scale to the thermometer, which converts the expansion of the mercury column into a digital signal.

In order not to connect converters every time, they came up with digital sensors. They also measure analog values, but they already have a built-in transmitter.

So, an electronic thermometer measures the same analog value as an ordinary thermometer – temperature. It has a built-in thermistor – an element whose resistance changes depending on temperature. The transducer inside the thermometer records the resistance and converts it into digital data. Then the thermometer displays these digital data on the display or transmits the data to the server or to the cloud.

Now digital sensors are almost always used in the IoT – it is more convenient to transfer data from them. Moreover, not ordinary, but smart sensors are becoming more and more popular.

What are smart sensors and why are they needed

One single sensor can only measure a specific physical quantity, such as moisture. But almost always business or manufacturing goals require more complex tasks than just measuring one metric.

For example: in production, we need to know the dew point – the temperature at which moisture from the air begins to fall out in the form of dew. To do this, you need to measure the humidity and temperature, calculate at what indicators the air will become too saturated with moisture, take into account the change in different indicators over time.

To solve this problem, you can install many sensors, link them, configure the system to collect data, and analyze it. Or you can just take a smart dew point sensor. It has all the sensors you need, as well as a processor that collects, processes, and analyzes data. As a result, such a smart sensor immediately transmits all the necessary information.

Some smart sensors even know how to make certain decisions. For example, a smoke detector can automatically trigger an alarm. He will transmit the alarm not information about smoke, but the command “turn on”, since he himself determined the level of smoke as dangerous.

It is very difficult to buy a dozen different sensors, connect them, and set up the processing logic. Therefore, now in the IoT, they mainly use smart sensors that are able to collect various indicators and transmit already primary processed data to the server.

How IIoT sensors differ from each other and how to choose the right ones

All sensors are different from each other – some can work in extreme conditions, others are more accurate, and others last longer. These differences affect the price – the more the sensor can do and the more it can withstand, the more it costs.

What parameters and in what situations are they important.

Sensor operating temperature – at what temperature the sensor can operate and display specific data. Typically, sensors can withstand temperatures from -20 ° C to 30 … 40 ° C. If the room temperature is low or high, for example, you need to measure something in the freezer, you will have to choose a more “hardy” sensor. Machine tools sometimes heat up or cool down to extreme temperatures too – this should be known and taken into account when choosing a sensor.

Housing protection – how protected the sensor is from moisture, dust and shock. Sensors on machine tools usually require serious protection from dust or pressure, but sensors in ordinary rooms can usually be installed unprotected.

Measurement accuracy – to what proportions the sensor fixes the value and what is its error. For example, for a temperature sensor in an office space, an accuracy of up to a degree is quite enough – it doesn’t matter if it’s 20 ° C there or 20.3 ° C. But in the temperature in production, accuracy to tenths or even hundredths can be important – then you need to choose sensors more precisely.

Measurement range. The minimum and maximum value that the sensor is able to capture. For example, if the measurement range of the thermometer is up to 50 ° С, 60 ° С it will show as 50 ° С – the sensor is simply incapable of more. The measurement range must be selected depending on the required accuracy and the values ​​that are encountered in your work.

Communication stability. How far is the sensor capable of transmitting data, whether the signal from it is afraid of interference and obstacles. The larger the enterprise, the more interference in the room, the more stable the sensor must be chosen. Otherwise, it turns out that a car will pass between the sensor and the receiver, and the data will be lost.

For reliable communication, there are modules that can store data and transmit it to other sensors with the same module. As a result, the data is saved, goes along the chain, and in any case gets to the server. These include, for example, ZigBee modules.

Uptime. How long will the sensor work without breakdowns subject to the conditions of its operation: temperature, humidity, pressure. Typically, long-term uptime is required if the sensor is to be installed in a hard-to-reach location.Size and weight. These indicators are important if the sensor needs to be installed in a confined space or on a small device. Generally, the smaller and lighter the sensor, the more expensive it is.