Oct 14, 2013The quantity of devices connected to the Internet exceeds the population of people on Earth. That's right—there are more devices tapping into the Internet than people on Earth to use them. Different sources predict that by 2020, wireless sensors and other types of wireless network nodes (such as actuators) will account for the majority (60 percent) of the total installed base of Internet of Things (IoT) devices.
Autonomous sensors will play a large part in those predictions. These sensors autonomously execute their functions in the environment in which they have been deployed. They are wireless, and their most distinctive characteristic is that they are self-powered while still being capable of monitoring the environment and transmitting data. Such devices range from simple detectors that trigger an alarm signal if the sensor passes a measurement threshold, to monitoring systems that collect data regarding different products or processes.
Autonomous sensors can potentially be deployed everywhere, though there are usually constraints on the power supply of such devices, so the communication technology must be carefully selected. Most of these sensors employ low-power-consumption wireless protocols, such as ZigBee or Bluetooth Low Energy (BLE). ZigBee is a protocol specifically developed for mesh communication—each sensing device can transmit data either directly to the acquisition module, or to another nearly ZigBee-based sensing device, which in turn will transmit both devices' data to the acquisition module. When they are deployed in urban areas, it is also typical to implement Wi-Fi-based devices, taking advantage of the Wi-Fi coverage associated with these places. A new, interesting trend with sensor devices connected to the IoT involves using smartphones to serve as a bridge between the sensors and the cloud. A force still to be reckoned with is that many citizens are eager to help their communities by installing smartphone applications that provide such a bridge to data, as long the information is to be used for something they believe in.
Power Supply of Autonomous Sensors
The power supply of these autonomous devices is the key. Even if research is continuously decreasing the power requirements of wireless sensors, the energy levels required are still too large a burden for the energy-harvesting technologies at hand. Power requirements from different communication protocols generally range from tens of milliamps to hundreds of milliamps. However, in most applications, there is no need to have the device continuously transmitting data, so the average power consumption is lower. The table below shows some general data comparing power requirements of the three wireless protocols widely used: Wi-Fi, ZigBee and Bluetooth Low Energy.
Please note that this comparison of power requirements is intended for the purpose of analyzing the impact on the power supply of autonomous sensor devices. There are, of course, multiple other variables to compare, such as communication range (Wi-Fi and ZigBee are much longer than Bluetooth), data rate (again, Wi-Fi and ZigBee are significantly better) and others.
In addition to wireless data transmission power consumption, we need to add the power consumption of the sensor itself. Fortunately there are many ultra-low-power sensors currently available on the market. For estimation purposes, we will agree on a sensor requiring some 50 microamperes (μA) to perform 10 measurements daily. Note that the total power consumption will range from hundreds of microamps to several dozen milliamps for a small amount of messages per day.
Natural Power Sources Available for Autonomous Sensors
Harvesting energy from the environment would allow these devices to operate on their own, virtually forever. The most-used natural power sources include photovoltaic, vibration, thermal and radio frequency (RF).
• Photovoltaic energy: Typically known as photovoltaic or solar cells, this power source can be a feasible solution, since the energy level harvested can be quite high, depending on the amount of space available. However, these devices require continuous exposure to sunlight, greatly decreasing their performance as the intensity of the sunlight is reduced.
• Vibration energy: There are multiple solutions for retrieving energy from environments in which continuous vibration is guaranteed—imagine a sensor device located in a train bogie (wheel assembly). Vibrational energy harvesting systems are designed for relatively stable conditions. For a selected frequency, the energy is correctly stored. However, for small frequency variations, efficiency dramatically decreases, thus reducing the amount of energy harvested over time.
• Thermal energy: The concept of taking advantage of a temperature gradient is not new. A thermoelectric device creates a voltage when there is a temperature gradient between the two ends of it. It is a very efficient system when the temperature difference is significant (the exterior and interior temperatures of an airplane in flight, for example).
• Radio frequency energy (RF): Signals from a variety of RF sources (television and radio transmissions, GSM signals from mobile phones and cell towers, Wi-Fi systems and so forth) can be captured by a sensor's antenna. The signals are then converted and conditioned to the desired output. RF is a constant energy source in every city of well-developed countries, but the amount of energy harvested is very limited. It requires long load times before reaching adequate energy storage levels for sensor nodes. Outside of cities, the RF source is generally very limited.
The conclusion is that no single energy source is reliable, unless environmental conditions are under control. Photovoltaic and thermal energy harvesting are very interesting in terms of the amount of energy that can be harvested, but the source must be very powerful in order for a sensor to draw a large level of energy—which is unlikely in most real-world applications. Because it takes a long time for an RF-harvesting device to store an adequate amount of energy to operate, RF energy can be suitable for sensors located near an RF source for which their activity is restricted to brief and very spaced communications.
However, a number of applications require a more reliable energy source before they can be implemented in the real world. Wireless battery-free sensors based on RFID are a very good fit in some of these cases.
Although it is not a natural source of energy, an RFID transmitter is a very interesting source to take into account for wireless sensors in the IoT scheme. A sensor that requires an RFID reader cannot be considered autonomous. However, this kind of device can be embedded at any point, and will always be ready to measure and transmit. As an example, passive (battery-free) RFID sensor tags can be embedded in concrete, inside piping systems and at many relatively inaccessible locations, and they will never require battery-change maintenance. The disadvantage is that they cannot measure or transmit when there is no nearby RFID reader to send RF power. Still, the advantages over natural sources are important:
A passive RFID sensor has an "on-demand" reliable source of energy. There is no dependency on environment conditions for the sensor to transmit the required data. It is also not affected by dark or hazardous locations, or by natural temperature changes.
Battery-free RFID sensors can be embedded inside different materials. This allows for the implementation of RFID sensors inside walls and pillars, or sealed within enclosures.
There is no doubt that energy harvesting is a growing industry, no matter the reasons—be they battery-free green policies or simple economics. Battery-free RFID sensors offer a way to enjoy reliable sensor data in your IoT scheme, without being affected by such unpredictable sources as sun, wind and temperature.
Mikel Choperena is the product development manager at Farsens, a manufacturer of ultra-low-power digital sensors and long-range passive ultrahigh-frequency (UHF) RFID sensors.