Why the Internet of Things Is Actually the ‘Internet of Sensors’
Ask the average person how their connected devices work, and chances are they will answer with “I just connect it with Bluetooth/Wi-Fi,” or something similar. Unless you are truly technologically inclined, the actual process that powers the Internet of Things is probably something akin to magic.
It doesn’t help matters that the “Internet of Things” is actually something of a misnomer. A more accurate description would be the “Internet of Sensors,” since the (IoT) quite literally couldn’t exist without smart sensors. Without sensors, your fitness tracker wouldn’t be able to count your steps, your smart thermostat wouldn’t know when to adjust the temperature up or down, and your smart refrigerator wouldn’t be able to tell you it’s time to buy more milk. Without sensors, companies wouldn’t be able to collect and record the data that is sent to the cloud, and develop better, more advanced products in the future.
Perhaps even more surprising to the average consumer is that the IoT, which often appears to be some of the most advanced technology available, actually relies on analog technology because of the need for sensors. While some have claimed that analog is old, outdated, or even dead, the fact is that it’s vital to the operation of pretty much any IoT device. The typical IoT devices uses analog sensors, at least to some extent, to collect the data that allows the device to actually operate. Data from the analog sensors is processed via an analog-to-digital converter (ADC), which translates the raw data from the sensors into digital form. That data is then aggregated, analyzed, and interpreted by the device’s onboard processors, which then determine the correct course of action.
The fact that the IoT relies on analog sensors doesn’t mean that the IoT is simple technology, though. In fact, as IoT devices become more advanced, so do the sensor configurations required to operate them.
Sensor Fusion and the IoT
Many of the most advanced IoT devices use multiple sensors, which require sensor fusion, which combines the data from multiple sensors to provide more information than is typically available from a single sensor. The thing is, this is often easier said than done. Sensor readings must be tightly calibrated and synchronized; otherwise, the algorithms used to analyze the data run the risk of becoming inaccurate and misleading.
Because inaccurate calculations can be detrimental to the operation of IoT devices, it’s becoming more important to use more integrated circuits that offer better signal conversion capabilities, through an op amp and other tools, and allow designers to build systems that can take accurate measurement of multiple sensors, and maintain the tight synchronicity required for sensor-fusion applications to work as they should.
Of course, the type of IoT application required creates different types of sensors, and different devices have a variety of requirements for accuracy, range, and bandwidth. For example, a weather monitoring s ystem that is only attempting to capture basic readings of temperature, humidity, and wind speed does not need the same converters and array complexity as a sensor fusion setup for complex physics experiments. That being said, since many IoT devices are controlled by multiple sensors, there is a need for more complex arrangements and signal-processing capabilities than in devices that have a single sensor.
The complexities of combining the worlds of analog sensors and digital processing are creating new challenges for designers. The IoT’s reliance on sensors means that microcontrollers are more analog intensive than ever before, and digital designers — the people responsible for developing new IoT technologies — now must develop interfaces that incorporate both analog and digital technologies.
That being said, digital sensors are an option, but for now they lack the accuracy offered by analog, and are considerably more expensive. Since most consumers now expect IoT capabilities from their devices, without a significant cost increase, analog sensors are still the dominant technology. Perhaps in the future, when digital sensors are more practical, the need for complex signal fusions to manage multiple sensors will diminish. Even when (or if) that happens though, the fact still remains: The IoT is really the Internet of Sensors, for without the sensors that create the “magic” consumers expect, nothing would work as it should.