To simplify the discussion, I’ll just use IoT because the same basic factors apply to all three. Being possessed of an engineer’s orderly mind, I’ll start with definitions.
A simple general definition of IoT is that it involves the communication of useful information among devices. For IIoT and Industry 4.0, those devices are in factories. The IIoT is a plant floor network of interconnected devices and machines. Industry 4.0 incorporates IIoT but expands that network to include higher-level systems for processing and analyzing data from the IIoT and using it for a wide array of functions, including looping back into the IIoT network.
In their most basic form, sensors provide raw data in the form of analog electrical signals. That simple fact poses a fundamental distinction: data vs. information.
What do I mean by that distinction? For example, a temperature sensor such as a thermistor has electrical resistance that varies with temperature. But that resistance is raw data — ohms — it has no meaning until ohms are somehow correlated to degrees. The simplest way to make that correlation is to measure the resistance with a bridge that converts resistance to voltage. That voltage can be applied to a measuring instrument that is calibrated to display volts as degrees. Data has thus been converted to information. But that voltage can also be used to control an electric heater or communicate the temperature of a motor to a predictive maintenance system.
Once the raw data has been converted into information, its uses are almost endless. That gets us back to IoT — the communication of useful information among devices.
To be used for IoT, the analog data supplied by a sensor has to be converted to digital form. This is done by electronics, typically an analog-to-digital converter (ADC). The rapid growth of IoT applications is causing a lot of work to be done to speed up the networks, enable them to handle ever larger amounts of data, and make them more power efficient.
One way of handling large amounts of data is to send it over the internet to external data centers — “the cloud” — for analysis and storage.
There are several problems with that. As IoT applications grow and get more complex, especially with the use of AI, the amount of data to be analyzed is increasing at an explosive rate. That places a serious strain on the available bandwidth and on the power used in the process. Cooling data centers is a major, energy-intensive task, so anything that can be done to limit the power load on the cloud would have an important role in limiting wasted energy.
Using the cloud also increases latency — the time taken between when the signal is sensed, digitized, transmitted, analyzed, and finally returned to be used in the real world.
Sending data out of a local facility to be analyzed or stored in the cloud also creates opportunities for the data to be corrupted, either accidentally or intentionally.
A solution to many of these problems is to share the workload. Enter the “smart” sensor.
The sensor becomes smart when it is embedded into a tiny package that includes electronics to do preprocessing. Electronics can do signal conditioning, analog-to-digital conversion, and wireless transmission of the data.
Even more, some sensors have AI capability embedded, so they can perform the initial steps of processing data into information right within the sensor package — at “the edge.” (See Edge Processing with Embedded Artificial Intelligence, Sensor Technology, June 2022.)
One problem with adding more functions to the sensors is that it requires more power, which is a serious problem for wireless sensors, and sensors for the IoT are typically wireless. You can hard wire sensors into local control loops, but that becomes very cumbersome once you want the data to be used for IoT or Industry 4.0 networks.
So smart sensors nowadays also include a variety of power-saving schemes. Much effort has been devoted to developing system-on-chip (SoC) architectures that reduce the power used by embedded sensor electronics. One approach is to do the initial processing with a low-power analog processing chip that decides when to wake up the more energy-intensive digital circuitry. (See Always-On Intelligent Sensing at Microwatt Levels, Sensor Technology, May 2023.)
The IoT World Is the Smart Sensor World
Smart sensors are vital to advanced driver-assistance systems (ADAS), autonomous vehicles, and EVs. (See Artificial Intelligence in Cars — Inside the Brains, Sensor Technology, March, 2022.)
They are helping revolutionize healthcare, including remote from-home monitoring, and giving healthcare providers real-time insights into the body that were never before possible. For example, a blood pressure sensor that can be placed at the tip of a very fine catheter or guidewire, and then used in remote locations inside the body, such as heart chambers, inter-cranial arteries, or even inside the kidneys, during critical surgical procedures. (See Smart Sensors Are Improving Medical Care, Sensor Technology, June 2021.)
In the industrial world, they enable advanced predictive maintenance by sensing gradual changes in machine behavior. (See The Many Benefits of Smart IIoT Connected Factories, Sensor Technology, March 2022.)
And those are just a few of the examples.