ArticlePDF Available

Abstract and Figures

In this work, we present autonomous active tags. The power sources of these active tags employ energy harvesting techniques, specifically, solar and mechanical techniques. The integration of these techniques, and the storage of the energy obtained with a supercapacitor, converts the active tag into an autonomous device. These tags work in a low power mode in which they dynamically adjust their radio communication capabilities. Such a configuration depends on the application. We tested the tags in a real environment with testing parameters to check the modules, meaning more wake-ups over a longer time. Under these conditions, the tags gather enough energy to autonomously maintain standby operation on a sunny day for ten hours. In conclusion, this autonomous active tag is a demonstration that the integration of energy harvesting techniques, supercapacitor storage and the management of low power modes for transceivers, microcontrollers, and memories creates a device without energy dependencies that only depends on the requirements, and can be used in many applications related to, for instance, smart homes, smart cities, smart cars, and connected forests.
Content may be subject to copyright.
applied
sciences
Article
Autonomous Active Tag Using Energy
Harvesting Strategies
Ana Carretero * , Santiago Real and Alvaro Araujo
B105 Electronic Systems Lab, Escuela Técnica Superior de Ingenieros de Telecomunicación, Universidad
Politécnica de Madrid, Avenida Complutense 30, 28040 Madrid, Spain; sreal@b105.upm.es (S.R.);
araujo@b105.upm.es (A.A.)
*Correspondence: anacp@b105.upm.es; Tel.: +34-91-0672-242 (ext. 72242)
Received: 31 May 2020; Accepted: 28 July 2020; Published: 30 July 2020


Abstract:
In this work, we present autonomous active tags. The power sources of these active tags
employ energy harvesting techniques, specifically, solar and mechanical techniques. The integration
of these techniques, and the storage of the energy obtained with a supercapacitor, converts the active
tag into an autonomous device. These tags work in a low power mode in which they dynamically
adjust their radio communication capabilities. Such a configuration depends on the application.
We tested the tags in a real environment with testing parameters to check the modules, meaning more
wake-ups over a longer time. Under these conditions, the tags gather enough energy to autonomously
maintain standby operation on a sunny day for ten hours. In conclusion, this autonomous active
tag is a demonstration that the integration of energy harvesting techniques, supercapacitor storage
and the management of low power modes for transceivers, microcontrollers, and memories creates a
device without energy dependencies that only depends on the requirements, and can be used in many
applications related to, for instance, smart homes, smart cities, smart cars, and connected forests.
Keywords:
autonomous; energy harvesting; active tag; supercapacitor; solar energy; mechanical energy;
low-power consumption
1. Introduction
The world of tags has evolved over the last few years. Passive tags appeared first, and are now
very widespread in our society, for example, in contactless cards or as anti-theft systems. They store
limited data and the reader has to be close to power and communicate with these passive tags.
These characteristics mean they have to be used in a reduced physical space with a restricted amount
of data. Then came active tags, which allowed communication over longer distances and greater
information storage capacities. This brought a paradigm change as all the data required for an
application can now be stored in the tag, instead of, for example, a remote server. Anyone with
permission can access the history of a tag without the need for a server. However, they need to
continuously power the elements required to perform these functions. Batteries are usually the way
this is achieved, but they have to be recharged regularly. Supercapacitors are currently being used to
replace them; although they have to be recharged like batteries, they have a longer life. The use of
energy harvesting techniques is now widespread to provide energy in this kind of scenarios in order to
convert the tag into an autonomous device.
Autonomous active tags are electronic devices that obtain, store, and transmit data from the
environment in which they are placed without any interaction of any user. The dierence between
them and passive tags is basically the internal source of power that active tags include to feed the
device, because passive tags do not need their own energy source to operate.
Appl. Sci. 2020,10, 5260; doi:10.3390/app10155260 www.mdpi.com/journal/applsci
Appl. Sci. 2020,10, 5260 2 of 21
The main energy harvesting techniques to consider for the autonomous requirement are: radio-
frequency, thermal, mechanical, and solar [
1
4
]. There are other energy harvesting techniques that
are being developed at the present time, such as fuel cells [
5
] and bio cells [
6
]. All these technologies
generate energy to be stored in batteries and supercapacitors. Integrating energy harvesting techniques
with storage elements in active tags turns them into autonomous active tags.
Due to the evolution to an autonomous and connected world, here we are making an eort to
develop a device adaptable to a great many applications that can establish a connection to other devices,
is wireless and uses the energy that nature provides as much as possible. We shall now study the
devices with dierent features and dierent harvesting techniques to obtain power.
Some systems use the radio-frequency technique, for example, as the energy source to monitor
when food has gone bad [
7
]. The system that manages the power and the data is at a distance of
6 m from the source antenna, or, as a monitor of temperature, at a distance of 50 cm, highlighting
the large antenna that is needed to receive the power [
8
]. This last has a feature to store energy in
a supercapacitor.
Other systems use the thermal technique to obtain energy; for example, implementing a bracelet
surrounded by heat sinks with a thermoelectric module (TEM) as a wearable for dierent applications
powered by a battery [
9
]. The mechanical technique is used in [
10
], where a large sized system is
developed using a piezoelectric combined with a triboelectric to obtain energy to power 50 LEDs
(Light-Emitting Diodes) or some electronic devices such as a calculator or a temperature sensor.
A great majority of the systems use solar energy as the technique to obtain power. One of these
solar systems power wireless sensor nodes to extend their life, enabling them to wake up for less
time, and increase the default transmission range [
11
]. The solar technique is used even in space,
as satellites recollect the energy, save it in a battery and use it for transmissions [
12
]. There is an animal
monitoring device that uses solar energy to know their behavior and follow their route using GPS
(Global Positioning System). The device stores the energy in a supercapacitor [
13
]. Other systems
mix more than one energy harvesting technique [
14
,
15
], such as solar and wind-based, in order
to get enough energy to meet the demand of remote sensors and small-scale portable electronics;
while solar and radio-frequency can be used to supply wireless sensor networks using dual storage
with a supercapacitor and a battery. There is even a system-on-chip that implements all the techniques
contemplated in this study to feed wireless sensor networks using a battery or supercapacitor [16].
Table 1shows an overview of the previous energy harvesting techniques, where the energy is
stored, and the application of the study.
The main contribution of this paper is to ensure that an active tag can work autonomously, using
power consumption management strategies and energy harvesting techniques to do so. The active tag
itself is composed of a memory, a microcontroller and a transceiver. The three components together
have the functionalities of saving, receiving, and transmitting data. In all these processes, the tag is
able to write data in a non-volatile memory unit, which means that the processing of the data can
be done in the tag. The objective is to gather enough energy to maintain the standby operation of
the device and support all the wireless communication. This is achieved thanks to the integration
of the energy harvesting techniques, the management of the recollected power and the low power
consumption techniques, which allowed the autonomous system to be created. In other words, the
creation of an autonomous active tag is based on the integration of dierent techniques from dierent
fields to achieve the objective. These tags can be used in isolated places without the interaction of
anybody as they recharge themselves and send the required information that the final application will
need. They stay on for as long as the application needs. This will be very useful, for example, to know
the state of trac signals without stainteraction on a road, by simply getting close to the tag with
a reader inside a car while driving. This is called an interaction infrastructure-vehicle, which is the
application selected for testing. Another example would be for a connected house, storing information
of the environmental parameters or habits.
Appl. Sci. 2020,10, 5260 3 of 21
Table 1. Current methods of energy harvesting techniques and their applications.
Information about the Authors of the Study Energy Harvesting
Method
Energy Storage
Method Application
Park, J.D. and Ren, Z. [5] Fuel cell - -
Catacuzzeno, L.; Orfei, F.; di Michele, A.; Sforna,
L.; Franciolini, F., and Gammaitoni, L. [6]Bio cell Supercapacitor -
Lam, M.B.; Dang, N.T.; Nguyen, T.-H., and
Chung, W.-Y. [7]
Radio Frequency technique
Supercapacitor Food monitoring
AAlex-Amor et al. [8]
Radio Frequency technique
Supercapacitor Monitoring temperature
Watson, T.C.; Vincent, J.N., and Lee, H. [9] Thermal technique Battery Bracelet as a wearable
Zhao, C. et al. [10] Mechanical technique - Electronic devices as a
calculator or LEDs
Mukherjee, M.; Shu, L.; Prasad, R.V.; Wang, D.,
and Hancke, G.P. [11]Solar technique Battery Extend the life of
wireless sensor networks
Boddu, P.; Kumar, B.S.; Ganapathy, K.; Suresh, T.,
and Kiran, P. [12]Solar technique Battery For satellite
transmissions
Loreti, P.; Catini, A.; de Luca, M.; Bracciale, L.;
Gentile, G., and di Natale, C. [13]Solar technique Supercapacitor Animal monitoring
Silva-Leon, J.; Cioncolini, A.; Nabawy, M.R.A.;
Revell, A., and Kennaugh, A. [14]
Solar and
wind-based techniques -
To meet the demand of
remote sensors and
small-scale portable
electronics;
Cao, D.; Jia, J.-R.; Xie, M.-J.; Lei, Y., and Li, W. [
15
]
Solar technique and Radio
Frequency technique
Battery or
supercapacitor
To supply wireless
sensor networks
la Rosa, R.; Livreri, P.; Trigona, C.; di Donato, L.,
and Sorbello, G. [16]
Radio Frequency, thermal,
photovoltaic or
vibrational techniques
Battery or
supercapacitor
Feed wireless sensor
networks
This study Solar technique and
mechanical technique Supercapacitor Road Maintenance
2. Materials and Methods
A complete autonomous active tag has been developed in this work. The architecture is based on
two modules, as is reflected in Figure 1: the power module and the processing module. This figure
shows the scheme of the dierent modules and submodules that make up the device, as well as the
flow of the normal operation.
Appl. Sci. 2020, 10, x FOR PEER REVIEW 4 of 21
Figure 1. Autonomous active tag architecture. Normal operation: energy is obtained from solar and
mechanical harvest submodules, which is stored in a supercapacitor and regulated to feed the diverse
elements of the processing module. This is indicated by simple arrows. Bold lines reflect the
communication of data between submodules. Dotted lines show the elements oriented to debugging
the device. The double dotted line between storage and battery means that the battery is connected to
the storage submodule.
2.1. Processing Module
The processing module has four submodules: the microcontroller, the memory, the transceiver
and the antenna, as can be seen above in Figure 1. Each one will be described in its own subsection.
The main function of this module is the control over the tag operation, which includes the tasks
of transmitting, receiving and saving data. All these tasks use different strategies to provide full
functionality with the minimum power consumption. These techniques are based on programming
the low power consumption modes of the microcontroller and the transceiver, as well as
synchronizing them in conjunction with the memory EEPROM (Electrically Erasable Programmable
Read-Only Memory), for a correct running. In this case, the effort to achieve the best performance is
important because it will lead to a greater lifetime of the device.
The first technique is the use of a low working voltage, as low voltages usually imply less
consumption and a longer lifetime. In this case, the lowest voltage that components allow is 1.8 volts.
The second design decision is to select the components of the tag with less current consumption
among those that fit in the other features. The third is to work in the 868 MHz Europe license-free
band, which has a lower power consumption and simplified protocols. The upper one of 2.4 GHz is
a more frequently used band than the 868 MHz band, so it has more interferences. In the case of the
lower license-free band of 433 MHz, the 868 MHz band has a better compromise between
communication range and binary rate.
The workflow of this module can be separated in two directions. The first starts in the
transceiver, which receives a package through the antenna. The package is then sent to the
microcontroller to be processed and, here, read or written in the EEPROM. The second follows the
inverse process: the microcontroller reads the EEPROM data, processes them to form a package,
which then arrives at the transceiver, which sends it to the reader that had initially requested the
data.
The system stays in low power mode until data is received. In our application, performing
transmissions with no prior request is not allowed because the data will not have a receiver, but this
is not a real limitation of the tag. The reason for that restriction is that this behavior would squander
energy and, therefore, shorten the system’s lifetime.
An Operating System has been mounted on the device in which the system has been developed.
The one used is FreeRTOS (Free Real-Time Operating System), a real-time operative system kernel
for embedded devices [17]. Including an operating system adds complexity to the system as well as
some overheads in the processing time and power consumption. However, this decision was taken
due to the potential to greatly expand the capability of the device; specifically, by providing a
Figure 1.
Autonomous active tag architecture. Normal operation: energy is obtained from solar
and mechanical harvest submodules, which is stored in a supercapacitor and regulated to feed the
diverse elements of the processing module. This is indicated by simple arrows. Bold lines reflect the
communication of data between submodules. Dotted lines show the elements oriented to debugging
the device. The double dotted line between storage and battery means that the battery is connected to
the storage submodule.
2.1. Processing Module
The processing module has four submodules: the microcontroller, the memory, the transceiver
and the antenna, as can be seen above in Figure 1. Each one will be described in its own subsection.
Appl. Sci. 2020,10, 5260 4 of 21
The main function of this module is the control over the tag operation, which includes the tasks
of transmitting, receiving and saving data. All these tasks use dierent strategies to provide full
functionality with the minimum power consumption. These techniques are based on programming
the low power consumption modes of the microcontroller and the transceiver, as well as synchronizing
them in conjunction with the memory EEPROM (Electrically Erasable Programmable Read-Only
Memory), for a correct running. In this case, the eort to achieve the best performance is important
because it will lead to a greater lifetime of the device.
The first technique is the use of a low working voltage, as low voltages usually imply less
consumption and a longer lifetime. In this case, the lowest voltage that components allow is 1.8 volts.
The second design decision is to select the components of the tag with less current consumption among
those that fit in the other features. The third is to work in the 868 MHz Europe license-free band,
which has a lower power consumption and simplified protocols. The upper one of 2.4 GHz is a more
frequently used band than the 868 MHz band, so it has more interferences. In the case of the lower
license-free band of 433 MHz, the 868 MHz band has a better compromise between communication
range and binary rate.
The workflow of this module can be separated in two directions. The first starts in the transceiver,
which receives a package through the antenna. The package is then sent to the microcontroller to
be processed and, here, read or written in the EEPROM. The second follows the inverse process:
the microcontroller reads the EEPROM data, processes them to form a package, which then arrives at
the transceiver, which sends it to the reader that had initially requested the data.
The system stays in low power mode until data is received. In our application, performing
transmissions with no prior request is not allowed because the data will not have a receiver, but this is
not a real limitation of the tag. The reason for that restriction is that this behavior would squander
energy and, therefore, shorten the system’s lifetime.
An Operating System has been mounted on the device in which the system has been developed.
The one used is FreeRTOS (Free Real-Time Operating System), a real-time operative system kernel
for embedded devices [
17
]. Including an operating system adds complexity to the system as well as
some overheads in the processing time and power consumption. However, this decision was taken
due to the potential to greatly expand the capability of the device; specifically, by providing a firmware
architecture that can enhance the software development speed through a hardware abstraction layer,
multitasking management and dynamic resource management [18].
2.1.1. Microcontroller
Our tag design includes an STM32L071C8 (M0+cortex), a low-cost microcontroller, which has
low current consumption in its maximum workload as well as built-in low power consumption modes.
The maximum working frequency of 32 MHz, 192 KB for flash, 6 KB for EEPROM, and 20 KB for RAM
(Random Access Memory) memory of the chip is enough to perform the tasks described in previous
sections, and to accommodate the operating system.
The microcontroller is connected to the transceiver and the memory via SPI (Serial Peripheral
Interface) bus, where it is the master and the others are slaves. This chip is responsible for every data
transaction, for turning oand for waking up both the transceiver and the memory. Accordingly,
it commands the other components and executes the process to follow.
The mode in which normal operation is developed is the active mode (default mode). Their
working frequency can be configured. The microcontroller integrates low consumption modes which
we will use in our tag in order to decrease the energy consumption and extend the lifetime. We use the
stop mode with RTC (Real-Time Clock), which works at 32 KHz, which retains the RAM, the register
contents and the real time clock.
The plan is to use the stop mode with RTC, one of the modes with lower power consumption,
the majority of the implemented duty cycle, which depends on the application, except for the moments
in which it is woken up by the transceiver or it needs to process and save data that have been received.
Appl. Sci. 2020,10, 5260 5 of 21
The working mode changes to active mode (default mode) working at 16 MHz to reduce consumption
without losing functionality because higher frequencies usually involve a higher consumption. This last
mode is the one used for processing receptions and transmissions.
There are also connections via a button and a LED for programmer control, and a UART (Universal
Asynchronous Receiver-Transmitter) for debugging purposes. The LED will not be assembled in the
final device because it would increase the power consumption of the tag.
2.1.2. Transceiver
We select the S2LP for its low power consumption and its radio configuration parameters.
This transceiver works in the following bands: 413–479 MHz, 452–527 MHz, 826–958 MHz,
and 904–1055 MHz. The band selected is the 868 MHz frequency, as we have explained in the
previous section, if we need to communicate with another tag or a tag reader. The distance achieved
is closely related to the adaptation stage. The best performance is traduced in a tradeobetween
communication range and binary rate.
The selected transceiver has a low power consumption mode called the low duty-cycle mode that
allows the average power consumption to be reduced during receive operations. This mode builds a
synchronized network where both transmitter and receiver can periodically switch to low power mode
to reduce the average power consumption. When the transceiver is in reception mode (RX), it waits
for specific bytes of the packet heading to continue reading the packet. After that, if a transmission is
required, transceiver changes to transmission mode (TX). The configuration of this duty-cycling low
power mode will depend on the final application of the device.
This transceiver has a data rate configuration by default of 38.4 Kbps and a maximum package
size of 128 Bytes in our tag for our application. The 2-FSK (Frequency-Shift Keying) is selected because
the data that we expect to arrive, which is explained in the description of the reception scenario in
Section 2.4.2, can be coded in this way. However, this modulation can be changed if the application
requires it.
The transceiver is connected to the microcontroller through their SPI buses. Thanks to that,
we can configure the transceiver with the characteristics that we prefer. Other features selected for our
application and which are very influential in the operation of the transceiver are:
the power of the output signal: this determines the distance at which the signal is detected. We have
chosen 10 dBm because this is the configuration that gives us the greatest communication range;
the sensitivity for detecting outward signals: in our case, this is around 109 dBm.
The antenna selected is external and its central frequency is 868 MHz. The connector is SMA
(SubMiniature A) and is welded to the board by one of the edges.
2.1.3. Memory
The EEPROM is the M95M02, which has a capacity of 2 Mbit and works with a frequency of
5 MHz. A low power consumption EEPROM has been chosen with enough capacity to give versatility
to the autonomous active tag.
The microcontroller manages the memory modes using an SPI bus communication and it has
read/write access. By default, this submodule is in standby mode, which is the lowest power
consumption mode that the memory has. When the microcontroller needs to perform a read/write
operation, it switches the memory mode to active and returns to standby once finished.
2.2. Power Module
This module provides power supply to the system and is responsible for collecting the energy
of the environment. The four submodules that form this module are: solar, mechanical, storage and
regulation, and debugger. All of them will be described in their corresponding subsections.
Appl. Sci. 2020,10, 5260 6 of 21
As described in Figure 2, our device has three energy sources: solar, mechanical, and electrical via
USB (Universal Serial Bus). These correspond, respectively, to the solar, mechanical and debugger
submodules. The collected energy is stored in a supercapacitor, since the batteries, which are the
main alternative, have the shortest life. In addition, supercapacitors have a higher power density and
faster charge time in comparison to batteries. The outputs of the energy sources are connected to the
supercapacitor using Schottky diodes; so, the current of one source follows the correct path and does
not interfere with the current of the other sources. After the diodes, the current of the dierent sources
is added by connecting the current lines before they reach the supercapacitor. Finally, the energy stored
in the supercapacitor is converted by a regulator into the 1.8 V needed by the processing module.
Appl. Sci. 2020, 10, x FOR PEER REVIEW 7 of 21
Figure 2. Power module architecture. Path 1 gets energy from the two energy harvesting modules
(solar and mechanical) to charge the supercapacitor and the battery (if connected) to power the
regulation module. In this path, the debugger (USB) is not connected and, usually, neither is the
battery. Path 2 gets energy from the debugger module to charge the supercapacitor and the battery
(if connected) to feed the regulation module. In this path, the debugger is connected and the battery
is not usually plugged in. The USB and battery are only connected for debugging cases.
2.2.1. Debugging Submodule
The debugger submodule powers the storage and regulation submodules to pass to the
processing module directly. The USB port is used in these cases, as it needs a previous regulation to
obtain the required 4.2 V output in both cases.
We select the BQ25606 charger, which allows an input of 5 volts and an output of 4.2 V. This
charger has two outputs: the one that powers the storage submodule, which is useful for charging
the supercapacitor quickly; and the one that powers the processing module after passing through the
regulator submodule, bypassing the storage submodule, which is useful for testing the operation
modes of the tag and programming it.
The use of this submodule was not conceived for actual application scenarios, but for testing
purposes and will be discarded in the final system. This is why the architecture shown in Figure 2
has two paths: one for collecting energy from the environment, which will be the usual working mode
of the tag; and another for the debugger (USB).
2.2.2. Solar Submodule
This submodule harvests energy radiated from the sun using a solar panel. In the preliminary
tests, two solar panels were tested but we selected the AM-5608CAR, with an estimated operating
time of 40 years. This was chosen following two constraints: to optimize the size of the tag developed
(the selected panel has a size of 60.10 mm × 41.30 mm × 1.80 mm) and the voltage range of the
supercapacitor for the sake of efficiency. The other panel used in the preliminary tests
(MONOCRYSTL SOLAR CELL 313070004) has a size of 70 mm × 55 mm × 1.5 mm and a voltage range
greater than the one needed.
Figure 2.
Power module architecture. Path 1 gets energy from the two energy harvesting modules
(solar and mechanical) to charge the supercapacitor and the battery (if connected) to power the
regulation module. In this path, the debugger (USB) is not connected and, usually, neither is the
battery. Path 2 gets energy from the debugger module to charge the supercapacitor and the battery
(if connected) to feed the regulation module. In this path, the debugger is connected and the battery is
not usually plugged in. The USB and battery are only connected for debugging cases.
The voltage provided by the output of every energy input was set to 4.2 V. The main reason is that
batteries normally use this voltage to work and, as this is a prototype, there is an option to connect a
battery as an auxiliary source of power for debugging scenarios. Furthermore, 4.2 V is considered
a valid voltage because of the capacity of the supercapacitor, less voltage may imply reducing the
capacity and the time that the system is on, which is the opposite of what we are searching for.
To design this module, preliminary tests were developed in order to understand how techniques
work in a real environment. The thermal technique was also evaluated in conjunction with the two
techniques finally selected, solar and mechanical. The three techniques were tested to charge a capacitor
and to turn on a LED. In the case of thermal technique, this was checked with two Peltier modules
and thermal paste, but this technique was discarded due to the low voltage obtained in the capacitor.
An instantaneous peak of voltage of 10 mV was raised along four days, but, mostly, the voltage was
0 V. For the other two techniques, supercapacitor has voltage values that allow to turn on the LED,
continuously in the case of solar energy and occasionally in the case of mechanical energy.
Appl. Sci. 2020,10, 5260 7 of 21
2.2.1. Debugging Submodule
The debugger submodule powers the storage and regulation submodules to pass to the processing
module directly. The USB port is used in these cases, as it needs a previous regulation to obtain the
required 4.2 V output in both cases.
We select the BQ25606 charger, which allows an input of 5 volts and an output of 4.2 V. This charger
has two outputs: the one that powers the storage submodule, which is useful for charging the
supercapacitor quickly; and the one that powers the processing module after passing through the
regulator submodule, bypassing the storage submodule, which is useful for testing the operation
modes of the tag and programming it.
The use of this submodule was not conceived for actual application scenarios, but for testing
purposes and will be discarded in the final system. This is why the architecture shown in Figure 2has
two paths: one for collecting energy from the environment, which will be the usual working mode of
the tag; and another for the debugger (USB).
2.2.2. Solar Submodule
This submodule harvests energy radiated from the sun using a solar panel. In the preliminary
tests, two solar panels were tested but we selected the AM-5608CAR, with an estimated operating
time of 40 years. This was chosen following two constraints: to optimize the size of the tag developed
(the selected panel has a size of 60.10 mm
×
41.30 mm
×
1.80 mm) and the voltage range of the
supercapacitor for the sake of eciency. The other panel used in the preliminary tests (MONOCRYSTL
SOLAR CELL 313070004) has a size of 70 mm
×
55 mm
×
1.5 mm and a voltage range greater than the
one needed.
The output voltage of the solar panel is connected to a step-up DC/DC converter, which charges
the supercapacitor and provides a constant charging voltage, as well as preventing damage to the
supercapacitor from high fluctuations. Although this strategy causes energy losses due to the eciency
of the converters, this eciency is considered enough for our application.
The chosen converter is the LTC3105, which is used in energy harvesting applications.
This converter meets the abovementioned specifications and, moreover, has good energy eciency
thanks to the Maximum Power Point Control (MPPC), which maximizes the energy that can be
extracted from any source of power. The output voltage was set to 4.2 volts, which is the charging
voltage of the supercapacitor.
2.2.3. Mechanical Submodule
This submodule obtains energy from vibrations. The energy is generated through a low-cost
cantilever-type vibration sensor based on a piezoelectric film (PVDF). In the preliminary tests,
three piezoelectric films were tested. One of them (MSP1006) generated a very low voltage in the
working tests due to it does not have a mass at one extreme, so, it was discarded before check it in the
real scenario. Between the other two, MSP1007 and MSP6914 (both with the mass at one extreme),
the first one reach and maintain a voltage of 50 mV during four days; the second one had an increased
progression during four days and, finally, it reached a voltage of 1.023 V. This is the main reason why
the 1005939-1 MSP6914 was chosen. Moreover, it has a physical size of 17.80 mm
×
7.00 mm
×
9.40 mm,
which fits in our tag design. This piezoelectric is frequently used to monitor continuous or impulsive
vibration in cars or infrastructures.
The sensor is then connected to a step-up DC/DC regulator with a smaller range of voltages than the
chip in the solar module, as this type of technique generates far less energy. In addition, the converter
is needed to obtain a constant charging output, as in the solar energy technique. The chosen device,
the ADP5091, includes a Maximum Power Point Tracking (MPPT) control feature that keeps the
input voltage ripple within a fixed range to maintain the power conversion stable. Another notable
characteristic is that it has two dierent outputs; one that we have dedicated to the regulator submodule
Appl. Sci. 2020,10, 5260 8 of 21
to power the processing module directly (called SYS), and the other to charge the supercapacitor (called
BAT), both with the same output value, 4.2 V.
2.2.4. Storage and Regulation Submodule
The storage section supports both a supercapacitor and an external battery. As commented
before, the use of a battery is conceived in this first prototype for actual application scenarios with
testing purposes.
The supercapacitor has to support a voltage of at least 4.2 V, which is the output voltage of the
submodules connected: solar, mechanical and debugger. We have reached a compromise between
lifetime, size and cost. Considering these three factors, we have chosen one that has an estimated
operating life of 10 years with 1 Farad of capacity, measurements of 17.00 mm
×
8.50 mm
×
12.00 mm,
and an aordable price. The maximum stored energy is 4.2 mWh, 15.12 J. However, it can easily
be replaced by another supercapacitor with a dierent capacity and similar measurements if a
specific application requires it. The space reserved for a battery connector is for the auxiliary battery
mentioned before.
In the regulation section, the input comes from the supercapacitor (and battery if connected),
where the necessary minimum voltage for the regulator to work is 2.5 V. The output voltage of the
regulator powers the processing module with a fixed value of 1.8 V and an eciency of around 87.5%.
2.3. Physical System and Package
In this section, the developed device is shown. The tag board (Figure 3a) has a size of
47.5 mm ×51.1 mm, including the supercapacitor that extends it by 12 mm in height.
Appl. Sci. 2020, 10, x FOR PEER REVIEW 9 of 21
divides them has an opening in the middle for the wires that connect the solar panel to the board.
There are two more openings, one for the piezoelectric component and the other for the antenna.
(a)
(b)
Figure 3. (a) Tag board; (b) Fully assembled autonomous active tag.
2.4. Application and Scenarios
There are many possible scenarios that could be used to test this autonomous active tag, but we
decided to validate the design of our device by saving information in the tag related to road
maintenance, thanks to our collaboration with a company specialized in this topic. The collaborative
project with the company consists of developing intelligent tags (our autonomous active tag) for the
identification, monitoring and traceability of road materials.
We tested our system in a real environment without a battery. The specific application consists
of attaching the device to a road sign with medium/low traffic running at a velocity of 50–80 km per
hour, which means that it is exposed to weather conditions. In addition, this application can provide
us with insights into its potential application in autonomous driving communications vehicle to
vehicle (V2V) and vehicle to infrastructure (V2I), etc., which are topics with a great importance these
days. These applications make the idea of autonomous active tags necessary as it is impossible to
develop a passive tag for these conditions, where there is no close communication with the reader.
2.4.1. Power Module Scenario
The tag is placed with piezoelectric pointing to the road with the objective of obtaining as much
oscillations as possible with the highest effect. With this background, we are able to check both the
solar and the mechanical energy harvesting modules. Testing inside the lab does not give us the
opportunity to check how energy harvesting techniques really work.
Based on the application explained before, it is expected to obtain how both energy harvesting
modules work, the time the device needs to charge and discharge the supercapacitor, and if it is
capable of receiving and transmitting packages with the harvested energy. This energy is collected in
the supercapacitor.
Solar energy depends on the hours of daylight, although it is also important to know how much
energy is captured when the tag is in a shady place during the day. The supercapacitor charge has
been measured for several days to check these two scenarios.
Mechanical energy depends on the amount of energy generated by the oscillation of the
piezoelectric when a car passes next to the tag with different speeds. That energy was not measured
with any external system. We have checked that the energy obtained from the wind caused by
passing cars through the piezoelectric film, since the vibration associated to the ground, and
transferred to the traffic sign, cannot be measured with this device. The mechanical module was
tested with the tag in the same place. At this point, the solar panel is disconnected to isolate the
techniques and obtain a measure related only to the mechanical module. This module has two
Figure 3. (a) Tag board; (b) Fully assembled autonomous active tag.
The tag cover (Figure 3b) has a size of 57 mm
×
63 mm
×
19 mm. It is separated in two
compartments, one for the tag board (lower) and other for the solar panel (upper). The thin plane that
divides them has an opening in the middle for the wires that connect the solar panel to the board.
There are two more openings, one for the piezoelectric component and the other for the antenna.
2.4. Application and Scenarios
There are many possible scenarios that could be used to test this autonomous active tag, but we
decided to validate the design of our device by saving information in the tag related to road maintenance,
thanks to our collaboration with a company specialized in this topic. The collaborative project with the
company consists of developing intelligent tags (our autonomous active tag) for the identification,
monitoring and traceability of road materials.
Appl. Sci. 2020,10, 5260 9 of 21
We tested our system in a real environment without a battery. The specific application consists of
attaching the device to a road sign with medium/low trac running at a velocity of 50–80 km per hour,
which means that it is exposed to weather conditions. In addition, this application can provide us
with insights into its potential application in autonomous driving communications vehicle to vehicle
(V2V) and vehicle to infrastructure (V2I), etc., which are topics with a great importance these days.
These applications make the idea of autonomous active tags necessary as it is impossible to develop a
passive tag for these conditions, where there is no close communication with the reader.
2.4.1. Power Module Scenario
The tag is placed with piezoelectric pointing to the road with the objective of obtaining as much
oscillations as possible with the highest eect. With this background, we are able to check both the solar
and the mechanical energy harvesting modules. Testing inside the lab does not give us the opportunity
to check how energy harvesting techniques really work.
Based on the application explained before, it is expected to obtain how both energy harvesting
modules work, the time the device needs to charge and discharge the supercapacitor, and if it is
capable of receiving and transmitting packages with the harvested energy. This energy is collected in
the supercapacitor.
Solar energy depends on the hours of daylight, although it is also important to know how much
energy is captured when the tag is in a shady place during the day. The supercapacitor charge has
been measured for several days to check these two scenarios.
Mechanical energy depends on the amount of energy generated by the oscillation of the
piezoelectric when a car passes next to the tag with dierent speeds. That energy was not measured
with any external system. We have checked that the energy obtained from the wind caused by passing
cars through the piezoelectric film, since the vibration associated to the ground, and transferred to the
trac sign, cannot be measured with this device. The mechanical module was tested with the tag in
the same place. At this point, the solar panel is disconnected to isolate the techniques and obtain a
measure related only to the mechanical module. This module has two outputs, as has been previously
commented, BAT (connected to the supercapacitor) and SYS (connected to the regulator). They were
monitored for several days.
The tag starts to work when the voltage in the supercapacitor reaches 2.7 V, which goes against
the 2.5 V of the theoreticians; this is due to component tolerances. That fact has been checked thanks to
a LED that turns on when the system starts to work. With the energy of both techniques, the power
module has to power the processing module to stay on and to transmit and receive packages. In the
final version, the LED will not be assembled due to the question of consumption.
2.4.2. Processing Module Scenarios
There are dierent scenarios in processing module relating to the consumption of the module in
order to discover whether the operations modes of our application are supported by the power module.
The supercapacitor is the source of energy and it is where the current consumption is to be checked.
The device turns owhen the voltage in the supercapacitor is lower than 2.7 V. Power consumption
was measured with a moving coil ammeter and a B2902A Precision Source/Measure Unit [19].
Low power consumption scenario
This is the main tag mode as it is the most usual scenario along the tag’s lifetime. For that
reason, we optimized consumption in this mode, so it has the lowest power consumption with respect
to the others. There are periodic peaks in consumption when the transceiver changes from sleep
mode to reception mode. The duty-cycle is 50% for a period of 1 s, which is higher than what our
application needs, but is useful for evaluating the consumption of the reception and sleep modes.
The microcontroller stays in the stop mode with RTC. EEPROM stays in standby mode because no
read/write operation is used in this scenario.
Appl. Sci. 2020,10, 5260 10 of 21
In this scenario, the EEPROM, the transceiver and the microcontroller are in low power
consumption modes, as has been explained in the description of this test’s setup. We use the
name “low power mode” to refer to this configuration. The transceiver wakes up in reception
mode during 0.5 s every 1 s, where the transition switches between reception mode and sleep mode.
The expected consumption, according to their datasheets, are between 34.5
µ
A and 41.5
µ
A in sleep
mode and 7 mA in the peaks of reception just for the transceiver. The consumption in stop mode is
0.8 µA for the microcontroller.
Since it is the mode in which the tag stays the longest, it is important to know how much time it
can be on while the previously fully charged supercapacitor is discharging, and without any source
of power. In order to do so, the current consumption has to be measured with the objective of
optimizing it.
Transmission scenario
In this scenario, the microcontroller reads data from the EEPROM and uses the transceiver to
send them to a hypothetical reader. Tests are carried out for one datum and for all data available in the
EEPROM. This test allows the transmission power consumption to be isolated. The microcontroller
uses the active mode continuously, so we can measure the transmissions consumption. The transceiver
switches to transmission mode (TX) when the microcontroller wants to send a package, which occurs
every second in this scenario. The EEPROM switches to active mode when performing read operations
and stays in standby mode the rest of the time.
First, it sends one datum, which is formed by 2 bytes of payload, and then it sends a data package
of maximum length, which is formed by 128 bytes of payload, at 38.4 Kbps.
The theoretical current draw needed to complete the transmissions is of 11.5 mA for transmission
mode (TX) in the transceiver, 3 mA for the use of the EEPROM and in the order of 0.87 milliamperes
for the microcontroller. The data are transmitted periodically at a rate of 1 package per 1 s.
Reception scenario
In this scenario, the data reach the transceiver and then pass to the microcontroller, which processes
them to understand the purpose of the package. Following that, it uses the EEPROM to accomplish the
specified task. Then, the microcontroller sends a response containing the requested data or a received
acknowledgement (ACK). This is the configuration that best fits with our application case, because,
a transmission is performed after every reception. The microcontroller the uses active mode all the
time for processing the receptions and transmissions as it is the best way to measure the consumption
of package reception. The transceiver switches between reception, sleep and transmission modes.
The EEPROM also switches between standby and active modes.
The transceiver is configured to be in continuous reception mode throughout these tests. In order
to test this mode, there are four alternative commands that a reader can send to our tag: read one
datum from EEPROM, read all data from EEPROM, read data saved at a specific date, and store a
datum in the EEPROM. These commands explain the cases that are the most similar to the cases of our
application: the tag receives a command, processes it and, if required, transmits an answer. The value
in the supercapacitor for the measures is 4.2 V.
1.
“Read one” datum from EEPROM: a reader system requires the last datum saved in the tag.
The transceiver receives a package from a reader with a “read one” command and sends it to
the microcontroller, which performs a read operation in the EEPROM. Thereafter, the datum is
transmitted back to the reader.
2.
“Read all” data command from EEPROM: here there are three transactions, because the transceiver
is not able to send all the available data in one packet. The process is the same as for transmitting
all the data in transmission scenario for the first transaction of this command. This first transaction
sends all possible bytes. The second transaction sends the bytes that could not be sent in the
Appl. Sci. 2020,10, 5260 11 of 21
previous transmission. The last transaction is to confirm that the transmission is ended, and this
means that this is a package with the lowest size, 2 bytes.
3.
“Read date”: here, the tag stores information from dierent days. This command retrieves the
information in the tag associated to a specific date. The transceiver changes from reception mode
to sleep mode. EEPROM searches for the specific date, after which the date found reaches the
microcontroller again, and the latter sends the date to the transceiver.
4.
“Write”: here a reader wants to write a datum in the EEPROM of the tag. The process is as follows:
the transceiver detects a command, changes from reception to sleep, the command is processed
in the microcontroller, it reaches the EEPROM, and after an examination of where to save the
datum, it is written. Then, the microcontroller checks that everything was correct and changes
the transceiver to reception mode again. This command does not have an acknowledgement
transmission packet; if we want to know if it has been saved correctly, we have to send a “read
date” command or “read one” datum command.
3. Results
The results have been divided into two groups. The first is focused on the energy harvesting
module, while the second is related to the aggregated power consumption of the tag.
3.1. Energy Harvesting Module
We first check the solar energy and then the mechanical energy. The tag turns on when the voltage
in the supercapacitor reaches 2.7 V. Figure 4shows the time it takes for the solar module to charge the
supercapacitor on a sunny day. The vertical grey line in 30 min and the horizontal orange line in 2.7 V
indicate the time and voltage, respectfully, needed to turn on the device. This means that the device
must be exposed to the sun for 30 min before it turns on because this is the time needed to reach 2.7 V.
The highest value for the supercapacitor is 4.2 volts, which is maintained after being reached.
Appl. Sci. 2020, 10, x FOR PEER REVIEW 12 of 21
orange line is the threshold at which the system is on, which is checked when the LED turns on and,
as is already known, happens at 2.7 V. The blue line is a square curve that represents when the system
turns on with “1”, and when system turns off with “0”.
Figure 4. Voltage in the capacitor with solar energy (blue trace). Thresholds refer to the working limit.
The cross between the vertical line (grey trace, 30 minutes) and the horizontal line (orange trace, 2.7
V) is when the system starts to work.
Figure 5. Voltage of supercapacitor with tag in a shady place for solar panel. There are four images
that represent how the voltage changes in the supercapacitor for different days represented by the
yellow line. (a) Shaded conditions during the night between 23 and 24 July, 2019. The voltage drops
because there is no energy source. System is off. (b) Shaded conditions during 8 August 2019 at noon.
The voltage rises but it starts at low value, so this increase is not enough to turn on the system. System
is off. (c) Shaded conditions during 9 August, 2019 at noon as (b), but for more hours. The voltage
rises and the start point of the voltage is close to the threshold that turns on the system, so the system
turns on; however, when system is on it needs more energy, which explains why it always oscillates
Figure 4.
Voltage in the capacitor with solar energy (blue trace). Thresholds refer to the working limit.
The cross between the vertical line (grey trace, 30 minutes) and the horizontal line (orange trace, 2.7 V)
is when the system starts to work.
Appl. Sci. 2020,10, 5260 12 of 21
There are four graphics in Figure 5, which show the drop of voltage throughout the day in a shady
place for several days. The first shows the drop during the night, the next two show the drop during
some hours with daylight, and the last is a zoom of the curve of the last picture. The horizontal orange
line is the threshold at which the system is on, which is checked when the LED turns on and, as is
already known, happens at 2.7 V. The blue line is a square curve that represents when the system turns
on with “1”, and when system turns owith “0”.
Appl. Sci. 2020, 10, x FOR PEER REVIEW 12 of 21
orange line is the threshold at which the system is on, which is checked when the LED turns on and,
as is already known, happens at 2.7 V. The blue line is a square curve that represents when the system
turns on with “1”, and when system turns off with “0”.
Figure 4. Voltage in the capacitor with solar energy (blue trace). Thresholds refer to the working limit.
The cross between the vertical line (grey trace, 30 minutes) and the horizontal line (orange trace, 2.7
V) is when the system starts to work.
Figure 5. Voltage of supercapacitor with tag in a shady place for solar panel. There are four images
that represent how the voltage changes in the supercapacitor for different days represented by the
yellow line. (a) Shaded conditions during the night between 23 and 24 July, 2019. The voltage drops
because there is no energy source. System is off. (b) Shaded conditions during 8 August 2019 at noon.
The voltage rises but it starts at low value, so this increase is not enough to turn on the system. System
is off. (c) Shaded conditions during 9 August, 2019 at noon as (b), but for more hours. The voltage
rises and the start point of the voltage is close to the threshold that turns on the system, so the system
turns on; however, when system is on it needs more energy, which explains why it always oscillates
Figure 5.
Voltage of supercapacitor with tag in a shady place for solar panel. There are four images
that represent how the voltage changes in the supercapacitor for dierent days represented by the
yellow line. (
a
) Shaded conditions during the night between 23 and 24 July, 2019. The voltage drops
because there is no energy source. System is o. (
b
) Shaded conditions during 8 August 2019 at noon.
The voltage rises but it starts at low value, so this increase is not enough to turn on the system. System
is o. (
c
) Shaded conditions during 9 August, 2019 at noon as (
b
), but for more hours. The voltage rises
and the start point of the voltage is close to the threshold that turns on the system, so the system turns
on; however, when system is on it needs more energy, which explains why it always oscillates between
on/o. (
d
) Reflects the graphic (
c
), but with a delimited voltage range to see the variation in the voltage
and the zones in which the system is on.
The mechanical module was tested with the tag in the same place for several days, and the charge
measurements of the supercapacitor are reflected in Figure 6. SYS is represented by points because it
collected a lot of energy with every car that passed; but then, if no cars passed, the voltage dropped
quickly, so a continuous signal would, in this case, not be real. In BAT, the voltage is maintained for
longer because it is saved in the supercapacitor, so the line in the graphic is continuous. The threshold
voltage to turn on the tag is represented by the horizontal orange line.
Appl. Sci. 2020,10, 5260 13 of 21
Appl. Sci. 2020, 10, x FOR PEER REVIEW 13 of 21
between on/off. (d) Reflects the graphic (c), but with a delimited voltage range to see the variation in
the voltage and the zones in which the system is on.
The mechanical module was tested with the tag in the same place for several days, and the
charge measurements of the supercapacitor are reflected in Figure 6. SYS is represented by points
because it collected a lot of energy with every car that passed; but then, if no cars passed, the voltage
dropped quickly, so a continuous signal would, in this case, not be real. In BAT, the voltage is
maintained for longer because it is saved in the supercapacitor, so the line in the graphic is
continuous. The threshold voltage to turn on the tag is represented by the horizontal orange line.
Figure 6. The voltage in the supercapacitor collected with mechanical energy. For every time that we
take the measurement, there is a value for SYS (yellow points), a value for BAT (light green line) and
the state of the device, called System On/Off (blue line). The system never turns on because neither
SYS nor BAT reach the 2.7 V needed, which is marked by the orange line.
3.2. Complete Tag
Now, the results of testing the processing module with the supercapacitor as the source of
energy are shown. The device turns off when the voltage in the supercapacitor is 2.7 V, as is
commented below.
3.2.1. Low Power Consumption Scenario
The results of this test scenario are shown in Figure 7. In Figure 7a, the voltage drop over time
can be seen, and is represented until the working limit of 2.7 V. Figure 7b shows the current associated
to the low power mode and Figure 7c shows the transitions associated to the current when this
scenario is running.
With a voltage of 4.2 V in the supercapacitor, the consumption in reception mode is around 7
mA and the low power mode is around 31.85 µA, which, considering the theoretical values, are the
expected values. In terms of energy, taking into account that, in our test, the time in low power mode
is 0.5 s, the energy needed for the low power mode is 66.89 µJ. The peaks of reception use 14.7 mJ for
the other 0.5 s.
3.2.2. Transmission Scenario
Figure 6.
The voltage in the supercapacitor collected with mechanical energy. For every time that we
take the measurement, there is a value for SYS (yellow points), a value for BAT (light green line) and
the state of the device, called System On/O(blue line). The system never turns on because neither SYS
nor BAT reach the 2.7 V needed, which is marked by the orange line.
3.2. Complete Tag
Now, the results of testing the processing module with the supercapacitor as the source of energy
are shown. The device turns owhen the voltage in the supercapacitor is 2.7 V, as is commented below.
3.2.1. Low Power Consumption Scenario
The results of this test scenario are shown in Figure 7. In Figure 7a, the voltage drop over time can
be seen, and is represented until the working limit of 2.7 V. Figure 7b shows the current associated to
the low power mode and Figure 7c shows the transitions associated to the current when this scenario
is running.
Appl. Sci. 2020, 10, x FOR PEER REVIEW 14 of 21
Figure 7. Tag power consumption measurements in the low power consumption scenario. (a) The
voltage drop in supercapacitor without external source. (b) Extraction of the low power mode and
the average of the current consumption. Here it can be seen that this mode lasts the 0.5 s. (c) Extraction
of the first 20 s of graphic (a), but in this case, the current draw of the tag is represented. The peaks
represent when the transceiver changes to reception mode every 0.5 s.
The data are transmitted periodically at a rate of one package per second. Figure 8 displays the
currents for transmitting one datum and a data package with 4.2 V in the supercapacitor. The lower
consumption line pertains to the microcontroller due to its continuous active mode. The time needed
to perform the transmission of a data package is higher than that needed to perform a transmission
of one datum. In addition, the consumption for both graphics are similar for each element, as can be
seen in the values of the figure.
Figure 7.
Tag power consumption measurements in the low power consumption scenario. (
a
) The
voltage drop in supercapacitor without external source. (
b
) Extraction of the low power mode and the
average of the current consumption. Here it can be seen that this mode lasts the 0.5 s. (
c
) Extraction
of the first 20 s of graphic (
a
), but in this case, the current draw of the tag is represented. The peaks
represent when the transceiver changes to reception mode every 0.5 s.
Appl. Sci. 2020,10, 5260 14 of 21
With a voltage of 4.2 V in the supercapacitor, the consumption in reception mode is around
7 mA and the low power mode is around 31.85
µ
A, which, considering the theoretical values, are the
expected values. In terms of energy, taking into account that, in our test, the time in low power mode
is 0.5 s, the energy needed for the low power mode is 66.89 µJ. The peaks of reception use 14.7 mJ for
the other 0.5 s.
3.2.2. Transmission Scenario
The data are transmitted periodically at a rate of one package per second. Figure 8displays the
currents for transmitting one datum and a data package with 4.2 V in the supercapacitor. The lower
consumption line pertains to the microcontroller due to its continuous active mode. The time needed
to perform the transmission of a data package is higher than that needed to perform a transmission of
one datum. In addition, the consumption for both graphics are similar for each element, as can be seen
in the values of the figure.
Appl. Sci. 2020, 10, x FOR PEER REVIEW 15 of 21
Figure 8. Current draw by the tag in the transmission scenario at 4.2 V in the supercapacitor. The
upper graphic shows the transmission of one datum and the lower graphic the transmission of all
data. The numbers represent which element is the one that has the largest contribution to the
consumption of the tag in that interval in each case: 1 is for the transceiver, 2 is for the EEPROM and
3 is for the microcontroller. There is a value per number to represent the average consumption of each
zone in the two graphics.
The time needed to complete one of these transmissions is approximately 50 milliseconds for a
package with a simple datum and 450 milliseconds for a data package of maximum length, as can be
seen in Figure 8.
The associated energy to transmit one datum starts with the peak in the EEPROM. The value is
calculated with the sum of the EEPROM energy, 16.7 µJ, the transceiver energy, 127.9 µJ, and the
microcontroller energy between the last two, 313.6 µJ. The total is 458.2 µJ for one datum.
In the case of all data it is calculated in the same way than for one datum. The values are: 272.4
µJ for the EEPROM, 1.17 mJ for the transceiver and 2.54 mJ for the microcontroller. The total is 3.98
mJ for all data.
3.2.3. Reception Scenario
In this scenario, the transceiver stays in reception mode and the microcontroller in the active
mode for all commands. The current consumption for the transceiver in reception mode is 7.6 mA
and for the EEPROM in active mode is 0.99 mA in all cases. The value in the supercapacitor for the
measurements is 4.2 V.
First, the “read one” datum from EEPROM, represented in the upper graphic of Figure 9. The
time required for the whole operation is approximately 100 milliseconds. The associated energy of
Figure 8.
Current draw by the tag in the transmission scenario at 4.2 V in the supercapacitor. The upper
graphic shows the transmission of one datum and the lower graphic the transmission of all data.
The numbers represent which element is the one that has the largest contribution to the consumption
of the tag in that interval in each case: 1 is for the transceiver, 2 is for the EEPROM and 3 is for the
microcontroller. There is a value per number to represent the average consumption of each zone in the
two graphics.
Appl. Sci. 2020,10, 5260 15 of 21
The time needed to complete one of these transmissions is approximately 50 milliseconds for a
package with a simple datum and 450 milliseconds for a data package of maximum length, as can be
seen in Figure 8.
The associated energy to transmit one datum starts with the peak in the EEPROM. The value
is calculated with the sum of the EEPROM energy, 16.7
µ
J, the transceiver energy, 127.9
µ
J, and the
microcontroller energy between the last two, 313.6 µJ. The total is 458.2 µJ for one datum.
In the case of all data it is calculated in the same way than for one datum. The values are: 272.4
µ
J
for the EEPROM, 1.17 mJ for the transceiver and 2.54 mJ for the microcontroller. The total is 3.98 mJ for
all data.
3.2.3. Reception Scenario
In this scenario, the transceiver stays in reception mode and the microcontroller in the active
mode for all commands. The current consumption for the transceiver in reception mode is 7.6 mA
and for the EEPROM in active mode is 0.99 mA in all cases. The value in the supercapacitor for the
measurements is 4.2 V.
First, the “read one” datum from EEPROM, represented in the upper graphic of Figure 9. The time
required for the whole operation is approximately 100 milliseconds. The associated energy of this
command, since the moment that transceiver changes their mode, is: 12.5
µ
J for the EEPROM, 268.8
µ
J
for the transceiver and 317.9 µJ for the microcontroller; a total of 599.2 µJ.
Appl. Sci. 2020, 10, x FOR PEER REVIEW 16 of 21
this command, since the moment that transceiver changes their mode, is: 12.5 µJ for the EEPROM,
268.8 µJ for the transceiver and 317.9 µJ for the microcontroller; a total of 599.2 µJ.
Figure 9. Current associated to reception mode at 4.2 V in the supercapacitor for the command read
one datum in the upper graphic and read all data in the lower graphic. The transceiver is in reception
mode for this test. As in the previous image, the numbers represent which element is the one that has
the largest contribution to the consumption of the tag in that interval in each case: 1 is for transceiver,
2 is for the EEPROM and 3 is for the microcontroller. The upper graphic shows the reception of one
datum and the confirmation of the transmission, the datum read is the last saved in EEPROM. In the
lower graphic, the reception of all possible data and transmission confirmation are shown. The
number 2 reflects the EEPROM aggregated consumption and the number 3 is the peak for the
transceiver transmission.
Second, the reception of the “read all data” command from EEPROM, as can be seen in the lower
graphic of Figure 9. The current consumption is the same as that previously commented for EEPROM,
around 0.99 mA, and for the transceiver in every transmission, around 10.6 mA, for the three that this
read all data command requires. The time spent in the complete command is around 2.2 s. The
associated energy to the command, considering the three packets needed to complete it, are: first
transmission: 311.75 µJ for the EEPROM, 2.54 mJ for the transceiver, and 1.37 mJ for the
microcontroller; second transmission: 228.7 µJ for the EEPROM, 1.96 mJ for the transceiver, and 1.1
mJ for the microcontroller; third transmission: 0 J for the EEPROM, 0.22 mJ for the transceiver, and
0.3 mJ for the microcontroller. The energy needed between transmissions are 19.79 mJ and 18.22 mJ.
The wasted energy for this command is 46.04 mJ.
Third, “read date”, Figure 10. The consumption is around 0.99 mA for EEPROM and 10.6 mA
for transmission, as in previous commands. The time for this command is around 300 milliseconds.
In this case, the associated energy, since the point that the transceiver changes the mode, as in the
other commands is: 636.2 µJ for the EEPROM, 623.3 µJ for the transceiver, and 467.7 µJ for the
microcontroller; energy of 1.73 mJ in total.
Figure 9.
Current associated to reception mode at 4.2 V in the supercapacitor for the command read one
datum in the upper graphic and read all data in the lower graphic. The transceiver is in reception mode
for this test. As in the previous image, the numbers represent which element is the one that has the
largest contribution to the consumption of the tag in that interval in each case: 1 is for transceiver, 2 is for
the EEPROM and 3 is for the microcontroller. The upper graphic shows the reception of one datum and
the confirmation of the transmission, the datum read is the last saved in EEPROM. In the lower graphic,
the reception of all possible data and transmission confirmation are shown. The number 2 reflects the
EEPROM aggregated consumption and the number 3 is the peak for the transceiver transmission.
Appl. Sci. 2020,10, 5260 16 of 21
Second, the reception of the “read all data” command from EEPROM, as can be seen in the
lower graphic of Figure 9. The current consumption is the same as that previously commented for
EEPROM, around 0.99 mA, and for the transceiver in every transmission, around 10.6 mA, for the
three that this read all data command requires. The time spent in the complete command is around
2.2 s. The associated energy to the command, considering the three packets needed to complete it,
are: first transmission: 311.75
µ
J for the EEPROM, 2.54 mJ for the transceiver, and 1.37 mJ for the
microcontroller; second transmission: 228.7
µ
J for the EEPROM, 1.96 mJ for the transceiver, and 1.1 mJ
for the microcontroller; third transmission: 0 J for the EEPROM, 0.22 mJ for the transceiver, and
0.3 mJ for the microcontroller. The energy needed between transmissions are 19.79 mJ and 18.22 mJ.
The wasted energy for this command is 46.04 mJ.
Third, “read date”, Figure 10. The consumption is around 0.99 mA for EEPROM and 10.6 mA
for transmission, as in previous commands. The time for this command is around 300 milliseconds.
In this case, the associated energy, since the point that the transceiver changes the mode, as in the other
commands is: 636.2
µ
J for the EEPROM, 623.3
µ
J for the transceiver, and 467.7
µ
J for the microcontroller;
energy of 1.73 mJ in total.
Appl. Sci. 2020, 10, x FOR PEER REVIEW 17 of 21
Finally, the last implemented command is “write”. The entire process is shown in Figure 11. The
total command time is around 157 milliseconds. The associated energy is 567.6 µJ in total, 70.7 µJ for
the EEPROM, and 496.9 µJ for the microcontroller.
Figure 10. Power consumption of the tag after receiving a “read date” command at 4.2 V in the
supercapacitor. As in other images, the numbers represent which element is the one that has the
largest contribution to the consumption of the tag in that interval in each case: 1 is for the transceiver,
2 is for the EEPROM, and 3 is for the microcontroller. The EEPROM is active longer than in other
commands since a concrete date has to be found in the whole memory. After that, the result is sent.
Figure 11. Power consumption of the tag after receiving a “write” command at 4.2 V in the
supercapacitor. As in other images, the numbers represent which element is the one that has the
largest contribution to the consumption of the tag in that interval in each case: 1 is for the transceiver,
2 is for the EEPROM, and 3 is for the microcontroller. The EEPROM is used twice, once to find where
to save the datum and a second time to save it. After that, the processing associated with the command
ends. The microcontroller remains active while the checks are being made and the data is saved.
4. Discussion
As shown in the results obtained in a real scenario, solar energy is dominant over mechanical,
since it obtains more energy in an environment that is always changing and is unpredictable. This
technique is the bottleneck of our device because the weather cannot be estimated accurately, so the
Figure 10.
Power consumption of the tag after receiving a “read date” command at 4.2 V in the
supercapacitor. As in other images, the numbers represent which element is the one that has the largest
contribution to the consumption of the tag in that interval in each case: 1 is for the transceiver, 2 is for
the EEPROM, and 3 is for the microcontroller. The EEPROM is active longer than in other commands
since a concrete date has to be found in the whole memory. After that, the result is sent.
Finally, the last implemented command is “write”. The entire process is shown in Figure 11.
The total command time is around 157 milliseconds. The associated energy is 567.6
µ
J in total, 70.7
µ
J
for the EEPROM, and 496.9 µJ for the microcontroller.
Appl. Sci. 2020,10, 5260 17 of 21
Appl. Sci. 2020, 10, x FOR PEER REVIEW 17 of 21
Finally, the last implemented command is “write”. The entire process is shown in Figure 11. The
total command time is around 157 milliseconds. The associated energy is 567.6 µJ in total, 70.7 µJ for
the EEPROM, and 496.9 µJ for the microcontroller.
Figure 10. Power consumption of the tag after receiving a “read date” command at 4.2 V in the
supercapacitor. As in other images, the numbers represent which element is the one that has the
largest contribution to the consumption of the tag in that interval in each case: 1 is for the transceiver,
2 is for the EEPROM, and 3 is for the microcontroller. The EEPROM is active longer than in other
commands since a concrete date has to be found in the whole memory. After that, the result is sent.
Figure 11. Power consumption of the tag after receiving a “write” command at 4.2 V in the
supercapacitor. As in other images, the numbers represent which element is the one that has the
largest contribution to the consumption of the tag in that interval in each case: 1 is for the transceiver,
2 is for the EEPROM, and 3 is for the microcontroller. The EEPROM is used twice, once to find where
to save the datum and a second time to save it. After that, the processing associated with the command
ends. The microcontroller remains active while the checks are being made and the data is saved.
4. Discussion
As shown in the results obtained in a real scenario, solar energy is dominant over mechanical,
since it obtains more energy in an environment that is always changing and is unpredictable. This
technique is the bottleneck of our device because the weather cannot be estimated accurately, so the
Figure 11.
Power consumption of the tag after receiving a “write” command at 4.2 V in the supercapacitor.
As in other images, the numbers represent which element is the one that has the largest contribution to
the consumption of the tag in that interval in each case: 1 is for the transceiver, 2 is for the EEPROM,
and 3 is for the microcontroller. The EEPROM is used twice, once to find where to save the datum and a
second time to save it. After that, the processing associated with the command ends. The microcontroller
remains active while the checks are being made and the data is saved.
4. Discussion
As shown in the results obtained in a real scenario, solar energy is dominant over mechanical, since it
obtains more energy in an environment that is always changing and is unpredictable. This technique is
the bottleneck of our device because the weather cannot be estimated accurately, so the amount of
energy obtained cannot be predicted, it is, that is the point with the most uncertainty. In the shady
zones, the device might not work continuously due to all the hours of sunlight the device is in the
shade. Consequently, for the final device, it is important to minimize or eliminate the hours in which
the device is in a shady zone. However, the limiting case of solar energy comes from the voltage drop
during the night, as can be checked in Figure 4. This means that if we want the device to stay on
continuously, it will be important to characterize those conditions in future implementations because
this application is related to an enclosed framework, where the sun is needed for a positive energy
balance. If the weather of the place where the tag is implemented is usually cloudy, or the days are
short during a great part of the year, this solution will not be optimal, although a previous study must
be done to corroborate this. These conditions can be considered as the constraints of the work since
the conditions limit the energy obtained to work. Focusing on the board, there is no concrete limit
for consumption, energy or size, the work pursuit the lowest power consumption, the most energy
obtained, and the smallest size (as a portability and environmental matter).
Nevertheless, for mechanical energy, it can be seen that the system is not capable of working
in those environmental conditions. The maximum values of collected mechanical energy (in SYS),
which are probably due to an increase in road trac, are not enough to turn on the system if this
technique were the only one working in the device. When a vehicle passes near the tag, the voltage in
the supercapacitor increases and then drops quickly. The average collected energy is not enough to
charge the supercapacitor. The same applies to the BAT output, although the voltage is maintained
longer for BAT, because this output is connected to the supercapacitor. So, the contribution of this
Appl. Sci. 2020,10, 5260 18 of 21
technique to the charge and running of the system is very low. A higher inflow of cars and with higher
speeds is necessary to allow the system to use only this technique as the main technology and to
consider an option to include it in future versions. In a final device, for this application, mechanical
energy will not be included. However, for other applications, it can be very interesting to consider this
energy. Consequently, whether to include it or not should be valued depending on the scenario.
As for the processing module, it is shown that the consumption is adequate for applications
such as the one at hand. The energy consumption peaks correspond to transmissions and receptions,
which can be sustained by the energy collected in the supercapacitor, as the system is still working after
these said peaks. Consequently, the device will remain on over time for whatever scenario. Staying in
one scenario or another is variable and depends on the application. Such new applications may require
a dierent supercapacitor, or only one of the two available energy harvesting modules, etc. Including
a battery can cause a malfunction due to the fact that it is not capable of handling transmission and
reception peaks as well as the supercapacitor. Furthermore, it requires a long time to charge and
decreases the lifetime of the system because it has fewer cycles. Therefore, a battery should not be
included in the next designs.
For this first prototype, in the low power consumption scenario of this application explained
above, with the supercapacitor fully charged and no external source of power, the tag is able to
function autonomously for 7 min, as can be checked in Figure 7a. When the tag is placed on the trac
sign, with the energy harvesting modules connected. An autonomy of 10 h is achieved, and this is a
parameter needed to characterize this device and really prove that it is autonomous. This is due mainly
to solar energy, which can recharge the supercapacitor after the peaks of reception that are acceptable
to the tag. However, it has to be considered that it depends on the environment and the application.
The device, for this application, will work during daylight hours where the dominant mode is the low
power consumption scenario with a duty cycling of 50%. However, the duty cycling can be reduced to
0.2%, which means that the consumption is 25 times less and thus increases the lifetime. This value for
duty cycling is considered the optimal value for the device because it is the lowest value for which the
transceiver is able to work correctly. Therefore, in the final application, this will be the value for the
reception mode. This tag prototype turns oduring the night, but the previous change in reception
values and the option to replace the supercapacitor with a larger one will give the tag full autonomy
for the whole day, although a study should be done. If there is enough energy, the scenarios work
regardless of the time of day.
The usability of this tag as a self-powered device can be guaranteed thanks to Figure 5, where the
device supports every peak of the transceiver and is able to continue working. Assuming the change
in the duty cycling and the solar source power, the autonomy and viability of using this device as the
final self-powered device will have to be checked again.
The abovementioned points highlight the need for a specific solar panel, piezoelectric,
and supercapacitor, as well as the configuration of the built-in low power consumption modes
according to the end application for an optimal optimization of the device. This also requires a prior
study of the conditions of the set-up location, under what conditions the application will be used and
which requirements the tag will need. This prior work will allow the maximum exploitation of the
energy harvesting submodules and the regulation and storage submodule to obtain the best performance
regarding energy and lifetime. Although no comparison has been done to check the optimization of the
processing module, we have made an eort to find the best configuration of the dierent components
in order to achieve the minimum consumption according to the manufacturer’s information.
This study is performed to cover the autonomous active tags applications, which have a mode
called low power consumption, where the consumption is as low as 31.85
µ
A (66.89
µ
J), with peaks of
7 milliamperes for reception (14.7 mJ), which is a good low consumption.
Taking into account the energy that the supercapacitor is available to store when is full charged
and looking at the energy requirements of the device in every scenario, it is clear that the energy balance
is positive. The command that needs more energy is “all data” in reception scenario due mainly to
Appl. Sci. 2020,10, 5260 19 of 21
the energy wasted between transactions, something that has to be checked for the final application.
Moreover, the reception peaks have a high energy consumption, this point supports the previous idea
of reducing the duty cycle.
The tag could be used in almost every environment initially considered, which turns this device
into a useful system. However, depending on the requirements of the working modes, the time it is
alive will be dierent. Furthermore, the consumption of the components could be reduced and the
supercapacitor can be replaced if the energy requirement of this application does not fit the one needed
for the new application, so the useful time would be longer.
The tag case prevents it being damaged by dust or water. The sensitivity of the device is the same
as all electronic circuits; in the presence of such extreme weather conditions as high temperatures
or humidity, over time, the tag can be damaged. Nevertheless, it is the task of the mechanical
encapsulation to prevent this from happening.
The complete tag only has installation costs, since maintenance is not contemplated. The design
of the tag has been conceived to give it a longer lifetime. However, in the case of incorrect operation,
the tag will be replaced because the installation is easy and our tag is scalable. This means that our
tags do not need any maintenance. Taking into account the materials and the cost of the board, one tag
costs approximately 50 euros, not including the labor costs. Mass production would reduce the cost of
the tag.
5. Conclusions
Autonomous active tags powered by a supercapacitor and solar and mechanical energy harvesting
modules, capable of working independently for storing, transmitting and receiving information are
viable for one application made from a project with a company. Consequently, this can be extrapolated
to other applications, such as those related to V2I or smart houses. The tag has been tested in an outdoor
scenario that fits our purposes and, given the results, we conclude that some parameters depend on
the application, which are related to the dierent modes in which the transceiver, the microcontroller
and the EEPROM can operate, and to the energy requirements that are basically based on a solar panel
and supercapacitor. In addition, the use of a supercapacitor instead of a battery gives the tag a longer
independent life.
There is a great uncertainty in the amount of energy that is available in every moment as this
energy depends on external factors as the sun or the cars. So, the dierent modules and the integration
of them have been developed to work in the best possible way to obtain the larger quantity of energy
and to use it to the tag has a greater autonomy.
As has been said before, possible applications are not only roads and trac signs but every
environmental situation, such as the roof or a window of a house, or even a car. In an environmental
situation, there are such variables as the temperature that can be saved to keep a record of the variations.
Perhaps this information could be sent to another device in the communication range or to a reader
that comes for the data. It could be the same inside a house, to monitor the temperature of dierent
rooms, or to automatically control and monitor some elements like the air conditioning, the doors or
even electrical appliances. All these are examples of what is now known as an IoT (Internet of Things)
node. However, in our case, it is not necessary to plug it into the node to get power or change the
battery when it is depleted.
On the other hand, scenarios where there is a heavy load would be dicult to support, even with
a great supercapacitor.
As there is no other tag with the features and functions presented in this work, we consider that
this solution is competitive and aordable for this application. The autonomy is defined, a parameter
that other studies do not usually contemplate, and which we consider to be an important value
to characterize the system, improve it and really know that it is autonomous. The integration of
the dierent techniques is a good way to improve the functionality and scope of this work for a
real environment.
Appl. Sci. 2020,10, 5260 20 of 21
Author Contributions:
Conceptualization, A.C., S.R., and A.A.; methodology, A.C. and S.R.; software, A.A.
and S.R.; validation, A.C.; formal analysis, A.C. and S.R.; investigation, A.C. and S.R.; resources, A.A.; data
curation, A.C. and S.R.; writing—original draft preparation, A.C.; writing—review and editing, A.A.; visualization,
A.C.; supervision, A.A.; project administration, A.A.; funding acquisition, A.A. All authors have read and agreed
to the published version of the manuscript.
Funding:
This research was partially funded by the Spanish Ministry of Science and Innovation, under CDTI
program (ITAGs Project).
Acknowledgments:
The authors also want to thank Francisco Tirado from B105 Electronic Systems Lab for his
comments, which greatly improved the manuscript.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
Newell, D.; Duy, M. Review of Power Conversion and Energy Management for Low-Power, Low-Voltage
Energy Harvesting Powered Wireless Sensors. IEEE Trans. Power Electron. 2019,34, 9794–9805. [CrossRef]
2.
Shaikh, F.K.; Zeadally, S. Energy harvesting in wireless sensor networks: A comprehensive review. Renew.
Sustain. Energy Rev. 2016,55, 1041–1054. [CrossRef]
3.
Vullers, R.; Schaijk, R.; Visser, H.; Penders, J.; Hoof, C. Energy Harvesting for Autonomous Wireless Sensor
Networks. IEEE Solid-State Circuits Mag. 2010,2, 29–38. [CrossRef]
4.
Yun, S.; Zhang, Y.; Xu, Q.; Liu, J.; Qin, Y. Recent advance in new-generation integrated devices for energy
harvesting and storage. Nano Energy 2019,60, 600–619. [CrossRef]
5.
Park, J.D.; Ren, Z. Hysteresis controller based maximum power point tracking energy harvesting system for
microbial fuel cells. J. Power Sources 2012,205, 151–156. [CrossRef]
6.
Catacuzzeno, L.; Orfei, F.; di Michele, A.; Sforna, L.; Franciolini, F.; Gammaitoni, L. Energy harvesting from a
bio cell. Nano Energy 2019,56, 823–827. [CrossRef]
7.
Lam, M.B.; Dang, N.T.; Nguyen, T.-H.; Chung, W.-Y. A Neural Network-Based Model of Radio Frequency
Energy Harvesting Characteristics in a Self-Powered Food Monitoring System. IEEE Sens. J.
2019
,19,
8813–8823. [CrossRef]
8.
Alex-Amor, A.; Palomares-Caballero,
Á
.; Fern
á
ndez-Gonz
á
lez, J.M.; Padilla, P.; Marcos, D.; Sierra-
Castañer, M.; Esteban, J. RF energy harvesting system based on an archimedean spiral antenna for low-power
sensor applications. Sensors (Switzerland) 2019,19, 1318. [CrossRef] [PubMed]
9.
Watson, T.C.; Vincent, J.N.; Lee, H. Eect of DC-DC voltage step-up converter impedance on thermoelectric
energy harvester system design strategy. Appl. Energy 2019,239, 898–907. [CrossRef]
10.
Zhao, C.; Zhang, Q.; Zhang, W.; Du, X.; Zhang, Y.; Gong, S.; Ren, K.; Sun, Q.; Wang, Z.L. Hybrid
piezo/triboelectric nanogenerator for highly ecient and stable rotation energy harvesting. Nano Energy
2019,57, 440–449. [CrossRef]
11.
Mukherjee, M.; Shu, L.; Prasad, R.V.; Wang, D.; Hancke, G.P. Sleep scheduling for unbalanced energy
harvesting in industrial wireless sensor networks. IEEE Commun. Mag. 2019,57, 108–115. [CrossRef]
12.
Boddu, P.; Kumar, B.S.; Ganapathy, K.; Suresh, T.; Kiran, P. Solar Energy Harvesting from Solar Power
Satellite. Ssrn Electron. J. 2019, 2356–2365. [CrossRef]
13.
Loreti, P.; Catini, A.; de Luca, M.; Bracciale, L.; Gentile, G.; di Natale, C. The design of an energy harvesting
wireless sensor node for tracking pink iguanas. Sensors (Switzerland) 2019,19, 985. [CrossRef] [PubMed]
14.
Silva-Leon, J.; Cioncolini, A.; Nabawy, M.R.A.; Revell, A.; Kennaugh, A. Simultaneous wind and solar energy
harvesting with inverted flags. Appl. Energy 2019,239, 846–858. [CrossRef]
15.
Cao, D.; Jia, J.-R.; Xie, M.-J.; Lei, Y.; Li, W. Hybrid Low Frequency Electromagnetic Field and Solar Energy
Harvesting Architecture for Self-Powered Wireless Sensor System. In Proceedings of the International
Conference on Wireless Algorithms, Systems, and Applications, Honolulu, HI, USA, 24–26 June 2019.
[CrossRef]
16.
la Rosa, R.; Livreri, P.; Trigona, C.; di Donato, L.; Sorbello, G. Strategies and techniques for powering wireless
sensor nodes through energy harvesting and wireless power transfer. Sensors (Switzerland)
2019
,19, 2660.
[CrossRef] [PubMed]
17.
FreeRTOS-Market leading RTOS (Real Time Operating System) for Embedded Systems with Internet of
Things Extensions. Available online: https://www.freertos.org/(accessed on 15 January 2020).
Appl. Sci. 2020,10, 5260 21 of 21
18.
Farooq, M.O.; Kunz, T. Operating systems for wireless sensor networks: A survey. Sensors
2011
,11, 5900–5930.
[CrossRef] [PubMed]
19.
B2902A Precision Source/Measure Unit, 2 ch, 100 fA, 210 V, 3 A DC/10.5 A Pulse|Keysight (formerly Agilent’s
Electronic Measurement). Available online: https://www.keysight.com/en/pd-1983585-pn-B2902A/precision-
source-measure-unit- 2-ch-100-fa-210-v-3-a-dc-105-a- pulse?cc=ES&lc=eng (accessed on 21 November 2019).
©
2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).
... These sensors are still difficult to develop due to the power and space limitations, accessibility problems, and the ubiquitous automotive need for reliability. All this points to the use of energy harvesting (EH) and wireless power transfer (WPT) where possible [3]- [6]. Studies already published have highlighted the need to develop methods and strategies to be adopted according to the type of application to be implemented. ...
Conference Paper
Full-text available
This paper presents three different techniques for efficiently powering an energy-autonomous wireless sensor (EAWS) through both energy harvesting (EH) and RF wireless power transfer (WPT). The aim of the paper is to provide effective strategies and techniques to reduce, as far as possible, the cost of wiring of the automotive production process due to the continuous and constant increase in the use of sensors. The techniques employ a highly integrated state-of-the-art, ultra-low power 2.5 µW system-on-chip (SoC) system, designed for multi-source RF wireless energy harvesting and power transfer and are designed with the goal of minimizing and, where possible, eliminating the costly maintenance required by conventional wireless sensors. Specific examples are reported that define both the aspects of convenience and the limits of use. Index Terms-automotive, energy harvesting, radio frequency, wireless power transfer, wireless battery charger, lithium ion battery, wireless sensor networks.
Article
Full-text available
The continuous development of internet of things (IoT) infrastructure and applications is paving the way for advanced and innovative ideas and solutions, some of which are pushing the limits of state-of-the-art technology. The increasing demand for Wireless Sensor Nodes (WSNs) able to collect and transmit data through wireless communication channels, while often positioned in locations that are difficult to access, is driving research into innovative solutions involving energy harvesting (EH) and wireless power transfer (WPT) to eventually allow battery-free sensor nodes. Due to the pervasiveness of radio frequency (RF) energy, RF EH and WPT are key technologies with the potential to power IoT devices and smart sensing architectures involving nodes that need to be wireless, maintenance free, and sufficiently low in cost to promote their use almost anywhere. This paper presents a state-of-the-art, ultra-low power 2.5 μ W highly integrated mixed signal system on chip (SoC), for multi-source energy harvesting and wireless power transfer. It introduces a novel architecture that integrates an ultra-low power intelligent power management, an RF to DC converter with very low power sensitivity and high power conversion efficiency (PCE), an Amplitude-Shift-Keying/Frequency-Shift-Keying (ASK/FSK) receiver and digital circuitry to achieve the advantage to cope, in a versatile way and with minimal use of external components, with the wide variety of energy sources and use cases. Diverse methods for powering Wireless Sensor Nodes through energy harvesting and wireless power transfer are implemented providing related system architectures and experimental results.
Article
Full-text available
This paper presents a radiofrequency (RF) energy harvesting system based on an ultrawideband Archimedean spiral antenna and a half-wave Cockcroft-Walton multiplier circuit. The antenna was proved to operate from 350 MHz to 16 GHz with an outstanding performance. With its use, radio spectrum measurements were carried out at the Telecommunication Engineering School (Universidad Politécnica de Madrid) to determine the power level of the ambient signals in two different scenarios: indoors and outdoors. Based on these measurements, a Cockcroft-Walton multiplier and a lumped element matching network are designed to operate at 800 MHz and 900 MHz frequency bands. To correct the frequency displacement in the circuit, a circuit model is presented that takes into account the different parasitic elements of the components and the PCB. With an input power of 0 dBm, the manufactured circuit shows a rectifying efficiency of 30%. Finally, a test is carried out with the full RF energy harvesting system to check its correct operation. Thus, the RF system is placed in front of a transmitting Vivaldi antenna at a distance of 50 cm. The storage capacitor has a charge of over 1.25 V, which is enough to run a temperature sensor placed as the load to be supplied. This demonstrates the validity of the RF energy harvesting system for low-power practical applications.
Article
Full-text available
The design of wireless sensor nodes for animal tracking is a multidisciplinary activity that presents several research challenges both from a technical and a biological point of view. A monitoring device has to be designed accounting for all system requirements including the specific characteristics of animals and environment. In this work we present some aspects of the design of a wireless sensor node to track and monitor the pink iguana of the Galápagos: a recently discovered species living in remote locations at the Galápagos Islands. The few individuals of this species live in a relatively small area that lacks of any available communication infrastructure. We present and discuss the energy harvesting architecture and the related energy management logic. We also discuss the impact of packaging on the sensor performance and the consequences of the limited available energy on the GPS tracking.
Article
Full-text available
Download here: https://authors.elsevier.com/a/1YXQm15eiew8HK Abstract This paper presents a novel wind/solar energy-harvesting device based on the inverted flag concept that combines flexible piezoelectric strips with flexible photovoltaic cells to simultaneously harvest both wind and solar energy. Three inverted flags built using off-theshelf components were experimentally investigated under controlled wind and illumination conditions to analyse their dynamics and electrical power generation capability. Our results provide an improved understanding of the dynamics of inverted flags that incorporate flexible piezoelectric strips and flexible solar panels, and indicate that the piezoelectric strips mainly increase the damping of the flags, while the solar panels act as an added mass. The power measurements show that the wind/solar energy-harvesting device proposed here is a viable concept that is capable of generating up to 3-4 mW of total power, enough to meet the demand of remote sensors and small-scale portable electronics, for wind speeds varying from 0 m/s (calm) to about 26 m/s (storm/whole gale) and 1.8 kLux constant light exposure, suggesting a rather diversified range of potential practical applications.
Chapter
The development of micro-energy harvesting technology provides a new energy solution for wireless sensor nodes (WSNs). Due to the intermittent power supplied by single environmental energy source, this paper proposes a hybrid energy harvesting architecture that harvest magnetic field (50–60 Hz) and solar energy simultaneously, which aims to provide a sustainable power supply for WSNs. Firstly, the design of free-standing “I-shaped” magnetic field transducer is introduced, which can harvest 0.17–0.46 mW underneath 700 A power transmission line. A further design of a rectifier and matching circuit is conducted and the maximum power point (MPP) of the hybrid energy harvesting circuit is about 60% of the open circuit voltage and the conversion efficiency reaches 61.68%. The experimental results show that the hybrid solar and “I-shaped” transducer can accomplish “cold start” operation of the power management unit (PMU) under magnetic flux density of 4.5 μT and light intensity of 200 lx, which will also provide a promising supply of energy for WSNs.
Article
Most radio frequency (RF) energy harvesters exhibit extremely nonlinear behavior during the harvesting phase, which renders the simulation of these circuits most challenging in design. This paper proposes a method of efficient modeling this nonlinearity based on empirically achieved characteristics of RF energy harvesting. A model using a multi-layer artificial neural network (ANN) is introduced to approximate this behavior. A harvester prototype that uses the far-field RF energy harvesting technique at 915 MHz is developed to verify the proposed model. The results highlight good agreement between the predicted and the observed outcomes of the harvester. With its flexibility, the proposed model can be further applied to characterize any nonlinear device. In addition, a self-powered system embedded with a fully-passive sensor device to monitor food storage profile is presented in the paper. An analysis of the power harvesting performance shows that the device can self-start its operation where the RF power is as low as -10.5 dBm, and the maximum achievable power conversion efficiency is 43%. To demonstrate the applicability of the proposed self-powered food monitoring system, an experiment is carried out using pork. Support vector machine (SVM) classifiers with two types of features are investigated to compare their performance of classifying three levels of food condition. As a result, the best classification accuracy of 95.7.
Article
Energy harvesting and storage devices, including lithium-ion batteries (LIBs), supercapacitors (SCs), nanogenerators (NGs), biofuel cells (BFCs), photodetectors (PDs), and solar cells, play a vital role in human daily life due to the possibility of replacing conventional energy from fossil fuels. However, these isolated devices only have limited performance and/or sole applicability, and cannot provide enough energy for application with long-term run and the ever-changing working positions. This suggests that it is urgent to develop the fine self-powered systems to meet the growing demand of energy for long-term use in different environment scenes. Developing integrated power pack, combining energy harvesting and storage, is an effective path to obtain a small size, light weight, high density and high reliability energy system. In this review, eight types of multifunctional integrated devices, such as LIB&SC, LIB&NG, BFC&NG, PD&BFC, SC&PD, SC&solar cells, NG&SC&solar cell, and LIB&solar cells, for energy harvesting and storage are reviewed in a broad sense, and a comprehensive summary of the recent development trends and highlights in the integrated device fields is given. Finally, the challenges and future outlooks for their successful commercialization are featured based on the recent advances and important findings.
Article
Conventional thermoelectric energy harvester design strategies often focus on improving the material property thermoelectric figure of merit (ZT); however, enhanced performance through improved materials cannot be achieved without optimized integration of subcomponents. The low voltage output from a thermoelectric module makes it challenging to design a practically usable wearable energy harvester. In addition to consideration of effective heat dissipation along with matching module geometry, a voltage needs to be boosted at a usable voltage value to utilize produced power. Using a DC to DC voltage step-up converter adds additional design complexity as the power cannot be harnessed at maximum power point of thermoelectric module due to input impedance requirement of the power conditioning circuit. Moreover, additional power loss occurs due to inherent voltage conversion inefficiency, which also depends on the harvesting voltage value. A recent design framework on wearable thermoelectric energy harvester assumed that a Maximum Power Point Tracking (MPPT) boost converter can be utilized for thermoelectric energy harvester systems, however it is unable to accurately determine the maximum power point during operation because of the transient nature of thermoelectric systems. Additionally, an MPPT circuit consumes power and adds complexity to the system design, which may not economically justify the use of such a circuit. This work proposes an encompassing thermoelectric system design framework for small scale energy harvesting using only state-of-the-art commercial products. Particularly, thermoelectric module geometry design is examined with the incorporation of a DC to DC voltage step-up converter without the use of MPPT. The framework can provide guidance for further development of subcomponents and materials, as well as system integration. An operational wearable energy harvester system was built using off-the-shelf components and demonstrated a usable power output which provides experimental evidence for the proposed design strategy.
Article
Energy harvesting from ambient energy sources has gained increased attention due to its advantage of less maintenance and for removing the dependency on batteries in IWSNs. However, due to the dynamic nature of ambient energy sources and the position of harvesting nodes, energy-harvesting is not always available, resulting in unbalanced energy-harvesting in IWSNs. Although, some battery operated nodes are used, the limited lifetime problem still exists due to the non-harvesting nodes. In this article, we propose a scheme that combines the advantages of energy-harvesting and sleep-scheduling in hybrid solar energy-harvesting IWSNs and non-harvesting nodes. We present a model of the harvesting-node using a three-state Markov chain. The proposed harvest-use-store type architecture aims to guarantee an energy-neutral condition to avoid energy harvesting nodes from early energy exhaustion. The proposed approach allows to wake up a few more non-harvesting nodes to handle network coverage and connectivity during less-energy-harvesting intervals. Similarly, non-harvesting nodes are allowed to sleep by increasing the default transmission range of the solar-harvesting nodes during higher energy harvesting intervals prolonging network lifetime.