IEEE Potentials Magazine - May/June 2018 - 13

For portable projects, one must consider
the run-time requirements and power demand
of the board/associated components
when selecting an energy source.

pcduino4 nano
The pcDuino4 Nano is the smallest
microcomputer featured in this article-two-thirds the size of the Raspberry Pi 3. The board is equipped with
an HDMI port, an ethernet port, three
USB host ports, and 40 GPIO pins.
There is also a USB-on-the-go port
allowing the board to behave as either
a host or a peripheral. It is the only
board to have a built-in microphone
and infrared receiver. Some examples
of its uses include the Android Smart
TV box, home automation systems,
and video surveillance robots.

sT Nucleo L476Rg
The ST Nucleo is similar in terms of
components to the Freescale and
Launchpad devices. One main differentiating feature is that the Nucleo
supports CANs, which means it can
communicate with other CANenabled devices without a mediator.
The ST Nucleo is compatible with
some Arduino-based shields, giving
it more functionality. Some examples include dc motor control and
GPS vehicle tracking.

c.H.I.p
The C.H.I.P. microcomputer houses
80 GPIO pins, a USB connector, a
battery connector, and built-in Bluetooth and Wi-Fi. Some of the GPIO
pins feature UART serial, I2C, and
complementary metal-oxide-semiconductor serial interface functionality. The company provides both
VGA and HDMI adapter shields for
additional connectivity. A single 3.7-V
lithium-polymer battery can power
the device, making it easily portable.
Some examples of its uses include
portable gaming devices, spectrum
analyzers, home Wi-Fi security, and
Wi-Fi routers.

Development board
decision examples
This section focuses on our thoughts
regarding the selection of a development board for each of our capstone projects.

project A
This engineering capstone project
involves processing real-time video

from vehicle-mounted cameras and
issuing alerts to the driver about
the roadside dangers detected by
computer-vision algorithms. Due to
the video processing nature of this
project, low-power microcontroller
boards can be ruled out, as they are
incapable of processing the video
data in real time as required for
generating real-time alerts.
Several of the computer-vision
algorithms are parallelizable and
could thus benefit from a multicore
processor. Of the reviewed boards,
only the Raspberry Pi 3 Model B and
the pcDuino4 Nano have multicore
processors. These two boards are
quite similar in specification-they
both have CSI camera connectors,
1-GB RAM, and run Linux from a
micro SD card. Even though the
pcDuino4 Nano sells for less than
the Raspberry Pi 3 Model B (US$25
compared to US$35), author Michael
Lescisin's capstone team opted for
the Raspberry Pi 3 Model B due to
the much larger maker community.
OpenCV has been used and is well
documented on the Raspberry Pi 3
Model B (Fig. 1), but the same level
of support cannot be found for the
pcDuino4 Nano.

peripherals was not needed, so
boards such as the Raspberry Pi,
BeagleBone, Galileo, and pcDuino
were not considered.
For the design to be wireless, the
MCU needed to support connectivity
technologies such as Bluetooth, WiFi, or radio frequency (RF). Another
important consideration was the
need for the board to be open source,
as the actual node itself is a customprinted circuit board that incorporates the MCU of the development
board, along with all its dependent
components. This limits the options
to the TI, NXP, and ST products. To
keep the complexity of the design
low, the MSP-EXP430FR5969 (Fig. 2)
was chosen.
This MCU proved to be a good
option for a wireless node. It's low
power consumption allowed it to run
continuously-transmit data over

project b
This engineering capstone project
involves collecting sensor data (such
as temperature, humidit y, a nd
ultraviolet intensity) from a wireless,
solar-powered node and transmitting it to a personal device for visualization a nd storage. The key
requirement for this project is the
node's power consumption. The
node must sustain itself for long
periods-upward of 200 days. As a
result, it was essential to choose the
lowest-powered microcontroller unit
(MCU) with the appropriate number
of GPIO pins. The use of an operating system or any visual/audio

Fig2 the mSp-EXp430FR5969 Launchpad was used in one of our projects.

IEEE PotEntIals

May/June 2018

■

Table of Contents for the Digital Edition of IEEE Potentials Magazine - May/June 2018