Instrumentation & Measurement Magazine 24-4 - 90

of figures and an uncertainty. The lack of the third part of this
information makes the numbers coming from the instrument
unsuitable for a decision.
Moreover, an even more dangerous effect could be the
excessive confidence in the " measurement results " when considering
them accurate because the UAVs embed accurate
sensors. It is clear, in fact, that the measurements taken by
means of complex systems, such as UAVs, are almost always
indirect. Consequently, the measurement uncertainty of the
overall system can be much higher than the sensor one.
In order to assess the uncertainty of a measurement provided
by an instrument, it is necessary to: identify the
uncertainty sources; define a model of the measurement procedure;
and apply the law of propagation of uncertainty
according to the model [9]. To perform these steps, it is necessary
to draw the measurement chain of the instrument and
identify the contributions to the uncertainty of each functional
block of the chain.
The main functional blocks of a conventional measurement
instrument [6], are: the sensor, the amplifier, the Analog
to Digital Converter (ADC), and the computer block, as shown
in Fig. 1. In particular, the sensor is able to convert the physical
quantity to be measured into an electrical signal. Then, the
amplifier is responsible for the amplification (or attenuation)
and filtering of the electrical signal provided by the sensor
with the aim of converting it to an analog signal suitable for
the next stages. The ADC converts to digital the analog signal
downstream from the amplifier. The computer can process the
data resulting from ADC and store the measurements into the
memory. The processed data are displayed to the user through
a Graphic User Interface (GUI). The main uncertainty sources
for these blocks are: the environmental conditions that affect
the sensor operations, such as temperature, humidity, etc.; the
thermal noise, non-linearity, etc. that affect the amplifier block;
and the quantization noise that occurs during the analog to
digital conversion step.
The UAV-based measurement instrument, with respect to
the conventional one, typically includes additional functional
blocks needed to describe the operation of the entire instrument,
as shown in Fig. 2. In particular, the instrument includes:
the Inertial Navigation System (INS), which provides the acceleration,
the angular rate and the altitude measurements
to the position estimation block; the GPS receiver, which provides
the position measurements; the position estimation
block that fuses the acceleration, angular rate, altitude and position
measurements and delivers to the onboard computer the
UAV attitude, in terms of pitch, roll and yaw angles, the UAV
speed along the three axes, the flight altitude and the UAV position;
the onboard computer, which is not only responsible for
processing the data provided by the measurement chain from
the sensor for the specific application, but also for communicating
with the ground control system and acquiring the data
provided by the position estimation block; the wireless interface,
that allows the communication between the onboard
computer and the ground control station, which exchanges
data related to the navigation, the data acquired by the sensor
for mission and the commands for driving the UAV during
the flight; the ground control station that allows the control of
the UAV during the flight and collects and processes the data
provided by it; and the data management system, which elaborates
the data provided by the ground control station (usually
in off-line) and displays the obtained post-processing results to
the user through the GUI.
As it was expected, being the functional block chain of a
UAV-based measurement instrument makes it more complex
than a conventional instrument, and the number of
uncertainty sources affecting the measurements grows (Fig.
2). Specifically, the uncertainty sources affecting the sensor for
mission are not only due to the environmental conditions, but
they consist of all the effects caused by the UAV navigation,
too, such as the frame vibration, the magnetic disturbances,
the propeller turbulences; and additional uncertainty sources
must be considered, which affect the position estimation block,
such as frame vibration, thermal noise, temperature, humidity,
and electromagnetic disturbances.
According to the above definitions of the main functional
blocks describing a generic UAV-based measurement instrument,
the main uncertainty sources affecting a specific
measurand can be identified. Furthermore, by knowing the
signal processing steps to obtain the indirect measurement
of the measurand, a mathematical model that describes the
measurement procedure and considers all of the identified uncertainty
sources can be carried out.
To give an example, referring to the car accident area surFig.
1. A conventional measurement instrument: its main functional blocks and its main uncertainty sources.
90
vey reported in [10], the signal processing steps to obtain
the indirect measurement of the measurand, in this case the
length, expect that the UAV estimates its position and attitude,
thanks to a GPS receiver and an inertial measurement
unit. Such information is
then combined with the
images coming from the
main camera sensors or
the LIDAR to produce a 3D
point cloud of the accident
scene, from which length
measurements are derived.
To determine the length
measurement uncertainty
sources, it has also to be
taken into account that
IEEE Instrumentation & Measurement Magazine
June 2021

Instrumentation & Measurement Magazine 24-4

Table of Contents for the Digital Edition of Instrumentation & Measurement Magazine 24-4