Smart Sensors
  • 18 September 2021
  • IT

The fourth industrial revolution, which is represented by Smart factories, was fueled by the Internet of Things (IoT) technology, and the internet of things technology is itself heavily dependent on microcontrollers and sensors and what is known today as a smart sensor. Various researchers have used the term 'smart sensor' in a variety of contexts, ranging from sensors with a few active devices to provide a more reliable interface to the sensor to improve the quality of the sensed signal, to integrated sensors with a sophisticated electronic circuit block including both digital and analog circuitry that helps transform the sensor. 

Before we talk about smart sensors, let us look at what a “not smart” sensor is first. a “not smart” or, in more technical terms, a base sensor is a device that senses something around it. It could be a smell, sound, light, or anything. To put the definition is a clear and scientific term, one can say a base sensor is a device that detects changes in physical properties and produces an electrical output in response to that change. For example, a microphone is a sensor that converts audible sound into an electronic signal. And a thermocouple is a base sensor that converts temperature into voltage where the hotter it gets, the more volts it outputs. Sensors do not work on their own and do not produce easy-to-read data which makes them a little too complicated to use in today’s plug-and-play world.

However, in recent years, a wide consensus has emerged, with a smart sensor being described as one that is capable of:

1. providing a digital output.

a smart sensor should be able to interface with any digital pin, which is usually a standard way. It could be considered as a Data-generating, data-storage, and data-processing technology that works in two states: positive state (5 volts or 3.3 volts) and non-positive state (0 volts). Any form of data can be represented in a digital, or binary, way and it is the way all modern electronic devices use to communicate.

2. communicating through a bidirectional digital bus.

A bidirectional bus is a standardized way of digital, or binary, communication, it being a bidirectional means it should allow both reading and writing on the bus which in itself a way, or protocol, of sending the data pre-configured on both ends. some examples of bidirectional busses that are used in smart sensors: I2C (Inter-Integrated Circuit) bus which consists of 2 wires, one for the communication clock and the other for data transfer. Another example is UART (Universal Asynchronous Transmitter and Receivers) which also consists of 2 lines: one for sending data and the other for receiving data.

3. being accessed through a specific address.

Each smart sensor on a certain digital bus should have an identification address to distinguish it from other smart devices or smart sensors on the same bus, for example, when using two I2C smart sensors, one for temperature and another for humidity, for instance, both are connected using the same 2 wires to the microcontroller or the processor but each is distinguished by their respective address, a typical I2C address, for example, is 7bits long allowing 128 different devices 

4. executing commands and logical functions.

A smart sensor should utilize the read functionality from the bidirectional bus to receive commands from the microcontroller or the processor. Said commands could be anything from turning on or off the sensor, implementing self-test algorithms on the sensor, getting diagnostics data, up to increase accuracy or sampling time.

A smart sensor is described as one that possesses these characteristics. Furthermore, the smart sensor should be able to perform functions such as secondary parameter compensation (e.g., temperature), failure prevention and control, self-testing, autocalibration, and other computationally intensive operations. But how is a smart sensor different from a base sensor? A smart sensor should have 3 main units onboard: digital control unit, signal processing unit, and external communication and bus interaction unit. These units will allow the smart sensor to meet the capability criteria discussed above. Let us discuss these in more depth.

Firstly, the Signal processing unit. Signals that are recorded by sensors are typically low in amplitude, that is how high and low does the voltage or current go whilst reading from the environment, However, combining interface electronics and signal processing circuitry at the smart sensor provides a range of advantages, such as signal amplification, impedance transformation, signal filtering and buffering, and multiplexing. All of which is done by the processor in base sensors. In many applications it is usually the signal amplitude that is low, amplification of sensor signals remains one of the most critical functions. Amplification of these signals at the smart sensor before transmission to the main processing unit not only improves the overall signal-to-noise ratio, reducing the impact of environmental noise, but also allows for maximum use of the dynamic range of an analog-digital converter for those sensors that have an ADC in the smart sensor module.

Secondly, digital control unit. Compatibility with digital control and microprocessor-based systems is one of the most important criteria for smart sensors. The vast majority of high-performance sensors should have a digital output that can be accessed through a digital bus. Following the digitization of sensor data, a variety of signal processing schemes may be used to compensate for a variety of errors and flaws. Offset cancellation, auto-calibration, self-testing, fault detection and correction, and linearity correction are some of these features. Although analog circuits can perform some of these functions, digital signal processing techniques are much simpler to implement using simple circuit techniques. Furthermore, some of these functions might be easier to incorporate using a remote host processor rather than at the sensor location.

And lastly, a bus interface unit. A smart sensor must be able to communicate with a higher-level controller that oversees the entire device. As a result, a portion of the smart sensor circuitry should be devoted to interacting with the bus to share data with the controller. There are two topics that are particularly important.

To begin, each smart sensor should be able to communicate with various buses and bus protocols. Secondly, perhaps even more critically, is the communication interface and its complexity. Over the communication bus, a variety of data, including addresses, calibration data, identification information, calculated data, and programming data initiated by the controller, can be exchanged between the sensor and the controller. The communication interface, in its most complex form, should be able to receive and transmit data over the bus at a reasonably high speed with not only the central controller but possibly other sensing units in a distributed system.

The main criteria for certain other applications, on the other hand, are durability, low lead count, and low noise. One such use is in the automotive industry. Various serial buses are being produced, with a focus on reducing the number of bus lines, possibly at the cost of speed.

In conclusion, Smart Sensors are a natural response to the modern industrial world and the Internet of Things awakening. Smart Sensors enable industrialists, engineers and makers to create a smarter, more inclusive, and ambient measurements that do not require post processing after measuring and can be interfaced with low-end microcontroller. The emergence of Smart Sensors allows people of different backgrounds to work with, otherwise complicated base sensors, advanced Smart Sensors without having to understand how a sensor exactly works. All one needs is a microcontroller and connect it to the smart sensor and that is it.