ArticlePDF Available

Abstract and Figures

This paper focuses on the development of a new logistic approach based on reliability and maintenance assessment, with the final aim of establishing a more efficient interval for the maintenance activities for Unmanned Aerial Vehicles (UAV). In the first part, we develop an architectural philosophy to obtain a more detailed reliability evaluation; then, we study the intrinsic reliability at the design stage in order to avoid severe critical issues in the UAV. In the second part, we compare different maintenance philosophies for UAVs and develop the concepts of preventive and corrective maintenance that consider the system subjected (until real “hard failure”) to partial performance degradation (“soft failure”). Finally, by evaluation of the uncertainty through the confidence interval, we determine the new soft failure limits, taking into account the general knowledge of the systems and subsystems in order to guarantee the proper preventive maintenance interval.
Content may be subject to copyright.
Sensors 2018, 18, 3171; doi:10.3390/s18093171
Reliability and Maintenance Analysis of Unmanned
Aerial Vehicles
Enrico Petritoli 1, Fabio Leccese 1,* and Lorenzo Ciani 2
1 Science Department, Università degli Studi “Roma Tre”, Via della Vasca Navale n. 84, 00146 Rome, Italy;
2 Department of Information Engineering, University of Florence, Via S. Marta n. 3, 50139 Florence, Italy;
* Correspondence:; Tel.: +39-06-5733-7347
This manuscript is extension version of the conference paper: Petritoli, E.; Leccese, F.; Ciani, L. Reliability
Degradation, Preventive and Corrective Maintenance of UAV Systems. In Proceedings of the 5th IEEE
International Workshop on Metrology for AeroSpace (MetroAeroSpace), Rome, Italy, 2022 June 2018.
Received: 25 July 2018; Accepted: 17 September 2018; Published: 19 September 2018
Abstract: This paper focuses on the development of a new logistic approach based on reliability and
maintenance assessment, with the final aim of establishing a more efficient interval for the
maintenance activities for Unmanned Aerial Vehicles (UAV). In the first part, we develop an
architectural philosophy to obtain a more detailed reliability evaluation; then, we study the intrinsic
reliability at the design stage in order to avoid severe critical issues in the UAV. In the second part,
we compare different maintenance philosophies for UAVs and develop the concepts of preventive
and corrective maintenance that consider the system subjected (until real hard failure) to partial
performance degradation (soft failure). Finally, by evaluation of the uncertainty through the
confidence interval, we determine the new soft failure limits, taking into account the general knowledge
of the systems and subsystems in order to guarantee the proper preventive maintenance interval.
Keywords: UAV; unmanned; aerial; vehicle; RAMS; reliability; maintainability; preventive;
corrective; maintenance; hard failure; soft failure; uncertainty; confidence interval
1. Introduction
The problem of reliability of UAVs, like problems of maintenance and safety, have become
extremely important in recent years: engines became more robust, avionics was improved, etc.
Despite this, the approach regarding the reliability of drones is still too fatalistic.
Obviously, by means of reliability analyses which are available nowadays, we consider that the
absence of a driver or person on board does not allow us to design and realize the UAV with less
stringent standards with respect to those used for airplanes. The commercial aviation failure rate is
about 1/105 flight hours, while for drones, it has been verified at about 1/103 flight hours, so a higher
magnitude of two orders. From a different point of view, sophisticated UAV systems have an overall
failure rate of 25%. The aim of the paper, which is an extended version of [1], is to provide new ideas
to increase the reliability of a drone optimizing maintenance activities. For this, we start from the
philosophy of apportioning the percentages of reliability assigned (on average) to each system (and
subsystem), trying to optimize them according to safety requirements. On the other hand, it is
necessary to optimize the time intervals (and consequently the costs) of maintenance, taking into
account that all critical systems must absolutely support preventive maintenance criteria: in these
cases, we are helped by the concepts of soft and hard failure [2,3].
Sensors 2018, 18, x FOR PEER REVIEW 2 of 16
1.1. Definitions
Here, we will introduce a series of definitions which will be used throughout the paper.
1.1.1. Reliability
Reliability is a dynamic concept which is applicable to many fields, i.e., is not only strictly
technical. For it, a possible definition could be expressed in terms of probability, and in particular, as
“the probability that a system, subsystem or part is able to perform its specific function in a pre-
established time and under pre-established conditions”. One of the most important reliability metrics
is represented by the Mean Time Between Failures (MTBF), expressed in hours of activity; the higher
the value, the more reliable the equipment. For a part or a single subsystem, the MTBF is often
expressed as the reciprocal of reliability. For instance, the MTBF gives information about the level of
unreliability, and it typically shows the number of failures of a piece of equipment over an established
time, i.e., 10,000 h [4].
The Failure In Time (FIT) rate of a device represents the number of expected failures in one
billion (109) device-hours of operation. This parameter is widely diffused in the semiconductor
industry [57] and international standards.
1.1.2. Availability
This parameter is extremely important for ‘ready to operate’ system. It measures the number of
times for which the system under study is available or ready with respect to the number of times in
which the system is required. Typically, this parameter is presented in form of a percentage, where
100% is the theoretical goal [811].
1.1.3. The Environment
According to “MIL-HDBK-217F2”, [12] (see Table 1), the following operative environment is
considered for the reliability prediction:
Table 1. Environment definition.
Class 1
Definition 1
Airborne, Uninhabited, Fighter
Environmentally uncontrolled areas, which cannot be inhabited
by an aircrew during flight. Environmental extremes of
pressure, temperature and shock may be severe.
1 Source: MIL-HDBK-217F2.
2. RAMS Assessment
The Reliability, Availability, Maintainability, and Safety (RAMS) assessment is an important study
in the development of UAVs. This kind of analysis is mandatory if you want to increase the reliability
of a drone, its availability, and to reduce repair and maintenance costs [13]. Once an architecture has
been chosen, the RAMS assessment is very useful to identify all the critical elements that could
increase the failure rate [14]. It also allows us to characterize all the most stressed (or undersized)
areas of the project. Furthermore, the reliability prediction, for example, makes it possible to decide
whether to duplicate a safety critical system or to put it in derated conditions, with great savings in
terms of weight and power consumption [1519]. A comparison between the well-known but always
efficient technique of redundancy and the improvement of reliability must consider important
remarks such as norms, costs, limitations of spaces, and so on. The reliability analysis helps us to
assess the value of failures [20]. For instance, if some failures of a specific component happen in a
wider system, the failure rate, preventively predicted, is useful to establish if the number of failures
which is adequate to the overall number of components present in the system. Alternatively, it can
individuate a particularly problematic section [2124]. Finally, this kind of evaluation can be used to
Sensors 2018, 18, x FOR PEER REVIEW 3 of 16
assess the probabilities of findable damage events in a FMECA (Failure Modes, Effects and Criticality
3. How Reliable Does a Drone Have to Be?
During a typical mission, some failures are more critical than others: the loss of longitudinal
stability, the loss of payload data, or the turning off of the position lights are not of the same criticality
level. Therefore, it is necessary first to establish various levels of increasing gravity associated with
the fault [25]. Moreover, according to the specific kind of mission and the specific kind of UAV, it is
necessary to subdivide failures into subcategories [26]. Then, for each scenario, suitability and
preventively forecasted, it is necessary to define a minimum acceptable level of reliability. Finally,
even for the aircraft, it is necessary to define the criteria for the level of reliability, in terms of a level
which is strictly linked to the type of failure [2731].
Catastrophic failures: for these kind of failures, a crash of the drone is certain while injuries or
even the death of persons on the ground is possible.
Severe failures: heavy damages are expected and the probability of repairing the drone is low.
Moderate failures: cause a moderate degradation of the drone’s functions, which could lead to
aborting the mission; however, they are not cause of severe damage.
Soft failures: cause light degradation of the drone’s functions, but do not lead to the cancellation
of the mission.
The intrinsic reliability of an apparatus (in our case, a drone) is the reliability studied a priori [32];
however, unfortunately, a reliability study is often realized after the design phase. This approach can
lead to many problems during the management of the project, because a reliability study highlights
a series of sensible points, and produces a series of recommendations that are usually forwarded to
the designers. These should allow them to carry all the necessary modifications to the project in order
to improve it [3338]. The perspective is completely different in the case of intrinsic reliability: in fact,
the knowledge of the failure distribution of a system gives rise to the possibility, directly during the
design stage, of taking specific action to reduce criticality and upgrade the critical parts or subsystems
in advance, thereby increasing the overall level of reliability [39]. This, surely, increases the level of
responsibility of the designers, but at the same time, decreases the risk of criticalities (also called
Single Point Failures or briefly SPF) that might happen in future studies [4043].
This way of seeing the design phase, i.e., in which a reliability study is effectively used to help
the project, means that the figure of the Quality Assurance Responsible is frequently present from
the beginning. Reliability analyses primarily aim to find the minimum limits for the requirements
that allow UAVs to have at least a magnitude of better reliable [44].
Another benefit of these analyses is that they help us to understand which components or parts
of a specific subsystem are the most unreliable, and which are the most critical to the system [4550].
In Figure 1 a failure rate comparison between five different aircrafts is shown: Northrop Grumman
RQ-4 Global Hawk in gray, PR-3 in orange, General Atomics RQ-1 Predator in blue, AAI RQ-2
Pioneer (for the drone category) in light blue and General Dynamics F-16 jet fighter in yellow.
Sensors 2018, 18, x FOR PEER REVIEW 4 of 16
Figure 1. Failure Rate vs. Flight hours for the F-16 and some common drones (figure extrapolated on
the data of: Barnard Microsystems Inc. et al. [5]).
4. Reliability Assessment Hierarchy
Considering all the possible main systems and subsystems that form an UAV, Figure 2 depicts
the UAV hierarchy of the reliability assessment, showing their failure distribution every 103 faults
Figure 2. The Hierarchy of the Reliability Assessment (every 103 system failures) for UAVs. The
structure is subdivided into main systems (yellow), subsystems (blue), all together representing the
UAV system (orange); (figure extrapolated on the data of: Barnard Microsystems Inc. et al. [5]).
The following subparagraphs show how the failures are categorized, taking into account the
function of each subsystem.
4.1. Ground Control System (GCS)
The Ground Control System, also called GCS, is the part with the highest maintainability of the
whole UAV system; this is because it is easily accessible at any time during the mission. It is mainly
composed of COTS (Commercial Off-The-Shelf) components with large inventories. This does not
mean that it is less safe; the fact that it is ground-based allows the introduction of a good percentage
of redundant systems with hot and cold stand-by configurations, reducing the off-line time nearly to
zero [53].
Failure Rate (for 105hours)
Flight Hours
Flight Vehicles Failure Rate Comparison
Pioneer PR-3 Global hawk F-16 Predator
UAV Sys (1000)
Control Sys
Power Plant
Navigation Sys
Electronic Sys
Sensors 2018, 18, x FOR PEER REVIEW 5 of 16
4.2. Mainframe
The UAV mainframe is by far the strongest part of the whole system; this is designed with great
attention by means of CAD systems, which allow the developers to study and evaluate the loads of
the structure a priori. In general, the mainframe is appropriately oversized; in fact, even if this leads
to extra weight, it is undoubtedly a small price to pay for a safer structural system. Generally, the
most common failures occur due to fatigue cycles, soldering brazing, or untreated rivets [54].
4.3. Power Plant
The power plant itself is a rather reliable mechanical system, even if the subsystems could show
some breakdowns. Especially in long-term missions, insufficient fuel vaporization or poor cooling
can lead to overheating of the engine, or fatigue that could eventually lead to failure [55,56].
4.4. Navigation System
This system is the most important part of the vehicle; it is characterized by a high failure rate
compared to other systems. Nevertheless, it has the highest number of hot redundant
subparts/subsystems. Therefore, due to the high level of electronic miniaturization, it is possible to
replicate a large number of its subsystems [57] making it intrinsically reliable. The Hot (standby)
redundancy is a method in which one system runs simultaneously with an identical primary system.
Upon failure of the primary system, the hot standby system immediately takes over, replacing the
primary system.
Moreover, due to the strong integration derived from the experience of the development of
automotive applications, the aerospace world enjoys highly-reliable electronic components for
navigation (as Inertial Navigation SystemINS and Global Positioning SystemGPS receivers). A
second benefit comes from their high computing capacity that allows a parallel computing
architecture, greatly increasing the overall reliability [58].
4.5. Electronic System
In the previous discussion, we have deliberately separated the electronic system from the
navigation system. Even from a purely mechanical point of view, it may not be so evident how they
are separated from a functional and philosophical point of view. In fact, in order to prevent possible
interference, the electronic system separates all the electronic circuits which are not closely related to
navigation, such as the power supply and conditioning, to manage the telecommunication system to
the outside (satellite communications, ground-vehicle data links, etc.). Even in this case, the
hypothesis of redundancy has to be avoided because, for example, the weight of the harness would
be excessively high for such a small vehicle [59].
4.6. Payload
The payload is not contained in its own conditioned bay inside the fuselage, but is placed outside
in a mobile turret inserted directly in the aerodynamic flow. The turret itself contains several electro-
optical sensors like a thermal imaging camera, a Low Light Level Television, a laser tracer etc. From
a mechanical point of view, the turret is gimballed, allows ± 90° elevation, and 360° of continuous
azimuth rotation; the system also contains the ancillary electronics of the sensors and the movement
The turret is thermostatically controlled to ensure optimal operation of the electronics and to
prevent freezing of the kinematic devices in high-altitude flight conditions. For these applications,
the electronics will be chosen with consideration of their intrinsic reliability and highest temperature
operative range. The geometry of the cases of the components must be chosen carefully, as the aircraft
is subjected to frequent and abrupt changes in altitude, and therefore, pressure, that could stress some
components. It’s important to note that to put all the equipment into a sealed and pressurized
container would be too burdensome from the point of view of weight.
Sensors 2018, 18, x FOR PEER REVIEW 6 of 16
5. Multiplexed Systems
Redundancy of the most critical systems does not always lead to an increase in the reliability of
the whole system. In fact, the redundancy of a quite unreliable system means increases in the total
failure rate. In other words, the overall safety of the system will be increased, but certainly not the
reliability. A higher failure rate brings an increase in expenses for spare parts and in person-hours,
increasing operating costs [11].
On the other hand, the duplication of the most critical or vital systems is not the ideal solution,
as it increases the cost and the complications of the system, so it is necessary to find another way to
improve reliability [60].
A classic case is that related to propulsion systems: many UAVs have only one engine, but this
is a highly reliable system, even if its loss compromises the entire mission. The installation of two
engines would seem to be an ideal solution because the loss of one of these would not compromise
the final mission. However, the duplication of a motor means the duplication of all ancillary systems.
This in turn leads to a decline in the system’s overall reliability [61].
The ideal solution is based on two keywords: oversize and derating. We will therefore choose an
engine with characteristics that exceed the UAV requirements, and will work under ordinary
operating conditions, i.e., derated, or in a very relaxed way. In this case, we will see that, while
remaining a single point failure according to FMECA analysis, the engine has a considerably higher
reliability, as it will work at less than 50% of its capability; this condition reduces the rate of failure
6. The UAV as a Complex Maintenance System
Now we consider an UAV as a complex aerial system composed by m subsystems (or subparts)
defined as J = {1, 2, …, m}, and consisting of lj components. At the component level, we can
continuously control and check the degradation of a defined collection of physical parameters. The
physical conditions degrade monotonically during use, and are restored by maintenance actions. For
each component or subpart i I, Xi(t) indicates the degradation trajectory in a fixed time interval t
[0, ). Soft failure can be defined as the ability of a component, part, subsystem or system to
continue its work even if with degraded performance, i.e., up to the point when its reduced
performance exceeds a specific fixed threshold Hi (with Xi(t) > Hi), called soft failure one. Typically,
components subjected to thermal stress or mechanical degradation are hit by soft failures.
When Xi(t) exceeds Hi, a soft failure happens between two maintenance points (n 1)τ and nτ.
This implies an action of corrective maintenance (CM), which has a specific cost (ciCM) on the critical
component. This action is executed in a fixed time called maintenance point , as shown in Figure 3.
Figure 3. Degradation threshold of a system with a cycle of corrective maintenance only.
The period between the occurrence of the soft failure point and the maintenance point is
defined as “soft failure period”. This period defines loss of quality in production or poorer
performance with a cost rate indicated with ciP [61].
Sensors 2018, 18, x FOR PEER REVIEW 7 of 16
6.1. Degradation Model for an UAV
In this section, starting from the UAV degradation model, we will arrive at defining limits and
probabilistic criteria to determine the optimal maintenance interval that does not exceed the
corrective maintenance threshold: maintenance that yields effect when the gradual damage is
intolerable by the agreed-upon operational standards.
The random coefficient model is used to evaluate the level of degradation for the ith component
for a time [0, ) in a cycle of single maintenance Φi = {ϕi,1, …, ϕi,Q }, Q , then a set of random
parameters Θi = {θi,1, , θi,V }, V following a normal (LaplaceGauss) distribution [62].
The probability that the degradation at time reaches the threshold before time is:
 
The calculation of this probability is necessary because, in the next discussion, we will introduce
a certain degradation profile and calculate the probability that this has to overcome the various
critical failure thresholds.
We consider a complex system with the following degradation path (this type of degradation
has been chosen because it is typical of this kind of complex systems):
where  and  then:
 
 
For a random variable,  , we evaluate the cumulative density function :
Now we evaluate the probability in which, between the two time points (n 1)τ and nτ, the
control limit Ci is reached:
 
that is equal to:
 
The soft failure threshold Hi is reached before time point only if it has satisfied the following
Moreover, assuming the degradation path as monotonic (typical of this kind complex systems),
we have: Ci < Hi with and .
6.2. Uncertainty of Degradation in Corrective Maintenance
If  we have two occurrences for the maintenance decision at time nτ:
preventive maintenance (PM) (see Figure 4a) and corrective maintenance (CM) (see Figure 4b)
according to:
The probability that a preventive maintenance happens at the specific time after the
degradation level of the ith component has reached the control limit  is [63]:
Sensors 2018, 18, x FOR PEER REVIEW 8 of 16
Figure 4. (a) Maintenance limit of preventive maintenance; (b) maintenance limit corrective
Now we consider the monotonic expression in the preventive maintenance: consider, for
example, what is happening around the monotone function immediately after the 3t maintenance
interval (see Figure 5a).
After the maintenance interval, since one of the returns from the field of intervention is the
degradation state of the systems and subsystems, we know exactly what the degradation status of
the systems and subsystems is. In other words, we can quantify as Xi(3t) as the status of the
probability at the moment we are studying.
Immediately after the “3t moment”, we have a view of the drift in time of the value: a band of
uncertainty affects the probabilistic function (supposedly monotonous). The uncertainty is due to the
capability of controlling the state of degradation of systems (and subsystems) limited by our
confidence interval (see Figure 5ac) in terms of knowledge of the complete system.
Figure 5. Uncertainty evaluation of corrective maintenance: the inspection point at 3t in detailed and
expanded as a confidence interval.
Sensors 2018, 18, x FOR PEER REVIEW 9 of 16
Now we consider X1, ..., Xn as samples of the subsystems degradation status of normal density after
the revision Xi(3t) with unknown mean m and variance σ2 (known) and sample average . We have:
 
Therefore, we have:
Explaining the second member (see Figure 6):
Figure 6. Confidence interval area considered for uncertainty.
We have for the confidence interval of level  for m:
It is necessary to treat the limitation of this method: the term  can never be lower
than Xi(3t). This is because the system, despite monotonic evolution in a more optimistic than
linearized way (Figure 5bgreen line), cannot, for logical reasons of entropy, improve over time, or
have a negative degradation. This condition is only theoretically possible, and is due to the uncertainty
of the state of knowledge of the system. Therefore, to restore the physical consistency of the
uncertainty evaluation, it is necessary to add the condition:
Our prediction of the state of health of our system cannot disregard the knowledge of the state
of the subsystems: their number and state influence the possible uncertainty of the value Xi(t).
6.3. The Thresholds of Preventive Maintenance
Considering the above calculations, when we want to evaluate the occurrence of the preventive
maintenance intervals, we need to consider the confidence interval in terms of knowledge of the
Reconsidering the confidence interval and the n cycle, the second part of the expression (9)
Obviously, the real problem happens when the threshold is exceeded:
Sensors 2018, 18, x FOR PEER REVIEW 10 of 16
And the new sof failure limit is:
Therefore, it is necessary to keep the confidence interval as narrow as possible.
Knowledge of the subsystems is therefore essential for evaluation: we risk calculating the total
reliability of the system without evaluating the total accuracy that, in the worst case, would lead to a
wrong assessment of preventive maintenance or an incorrect evaluation of preventive maintenance.
Obviously, this is an undesirable situation.
The condition for the threshold  is:
Now, the first member of the (9) becomes:
Now we can objectively quantify the level of accuracy needed to define the preventive
maintenance intervals.
It is still useful to specify the function from a graphical point of view. Now consider the upper
zone of Figure 4a in detail (see Figure 7):
Figure 7. Uncertainty evaluation of corrective maintenance: original interval area (red) and the
evaluation of the confidence interval (green).
From Figure 7a, the correlation between confidence and useful interval for preventive
maintenance is evident (in Figure 7b the detail is enlarged and the confidence intervals evidenced):
the lower the confidence, the higher the probability that corrective maintenance is necessary.
Sensors 2018, 18, x FOR PEER REVIEW 11 of 16
6.4. The Failure Rate Paradox
Now lets examine the real case of two completely different drones: a commercial drone and a
surveillance drone. Both have on board, as payload, a system of stabilized cameras: in our reliability
study, we will examine and compare only the systems and subsystems which they have in common.
Let us now compare the reliability of the average commercial drone: a reliability profile has been
created as a weighted average of the data present in our database, created through previous research.
This is compared to an average military drone created according to official sources [64].
Furthermore, the reliability of all subsystems is compared; it is immediately evident that the
distribution is different.
According to Table 2, it is absolutely evident that the military drone, due to its complexity, has
a reliability that is considerably inferior to that of a commercial drone which is certainly not built
with stringent parameters and requirements.
Table 2. Comparison between the reliability of a commercial and a military drone.
Commercial Drone (a)
System Description
λP System
FIT (F/106 hrs)
Ground Control System
Power plant
Navigation system
Electronic system
MTBF (RTotal) =
Military Drone (b)
System Description
λP System
FIT (F/106 hrs)
Ground Control System
Power plant
Navigation system
Electronic system
MTBF (RTotal) =
As is known, the commercial drone is composed entirely of COTS parts. Although there are not,
for example, MIL reliability-level electronic components, the reliability of commercial electronic
components is now extremely high, even those with plastic casing. This is also a consequence of the
continuous research in the automotive field, where the operating temperature range is quite high. All
of these aspects, combined with low construction complexity, lead to a high reliability level.
The military drone, on the other hand, enjoys a high degree of redundancy and a knowledge of
high quality components, but being an extremely sophisticated and complex product, it is heavily
penalized from the point of view of the reliability figure.
It must stated, however, that the capabilities of the latter compared to the commercial drone are
noteworthy: greater range, greater autonomy, higher payload, and resistance to soft failure. These
Sensors 2018, 18, x FOR PEER REVIEW 12 of 16
are all characteristics that are transparent to the calculation of reliability, and that then, eventually,
lead to the paradox.
In light of these considerations, we review, in Figure 8 (Figure 8a refers to “drone a” and Figure
8b refers to “drone b”), the previous confidence interval area considered for uncertainty in Figure 6:
Figure 8. Confidence interval area considered for uncertainty: commercial drone (blue) and military
Due to the good knowledge of the systems and subsystems of the military drone (hereafter
referred to as drone b), we can have the basis for a much wider Gaussian, and conversely, the value
of α.
Reviewing the interval in Figure 7, for the two different drones we have the following
uncertainties (see Figure 9):
Figure 9. Uncertainty evaluation of corrective maintenance: original interval area (red) and the
evaluation of the confidence interval (green).
Considering “Drone a”, although its reliability is considerably higher, the knowledge of the
components is lower. All this is reflected, from the analytical point of view, in the “shrinking” of the
green band (see Figure 9), or the limits of preventive and corrective maintenance which are very close
to each other. From the real point of view, this means that if we do not want to overcome the new
limit , it is necessary to slightly reduce interval t, with a consequent increase of
maintenance costs and a decrease in the general figure of availability.
The “Drone b”, (military) has, despite a bit of shearing by the α factor dropping a good margin,
an increase in the frequency of the maintenance cycle, which will always remain modest.
Sensors 2018, 18, x FOR PEER REVIEW 13 of 16
7. Conclusions
In this paper, the uncertainty in the choice of the preventive maintenance intervals with respect
to the soft failure threshold have been investigated, taking into account the reliability and safety
requirements for Unmanned Aerial Vehicles (UAV).
First, we examined the state of the art of the philosophy of the UAVs and the roles and
capabilities of operators. However, the increase of their use is strongly accompanied by higher failure
rates compared to conventional, manned airplanes.
Then, we correlated the reliability of the drones with the maintenance intervals: a higher failure
rate leads to very expensive repairs. In order to improve safety, the duplication of the troublesome
elements is not the only solution. Therefore, it is necessary to obtain the required reliability level by
also using high-quality, derated components, combined with a very detailed selection of a redundant
subsystem during the design phase.
The innovation of our paper passes first through the review of the optimization of the
probabilistic functions (under the conditions of a real case). Then, we find the optimal point of
maintenance. It will be necessary to take into account a very large number of variables for all systems
and subsystems in order to minimize uncertainty.
Finally, by evaluating uncertainty through the confidence interval, it is possible to accurately
determine the maintenance intervals in order to not exceed the new soft failure limit, that takes into
account the general knowledge of the systems and subsystems, and to remain always within the
preventive maintenance limit time (and budget).
Author Contributions: The three authors have made an equal contribution in all the sections of this article.
Funding: This research received no external funding.
Conflicts of Interest: The authors declare no conflicts of interest.
1. Petritoli, E.; Leccese, F.; Ciani, L. Reliability Degradation, Preventive and Corrective Maintenance of UAV
Systems. In Proceedings of the 2018 5th IEEE International Workshop on Metrology for AeroSpace
(MetroAeroSpace), Rome, Italy, 2022 June 2018, pp. 430434. doi:10.1109/MetroAeroSpace.2018.8453629.
2. Hobbs, A.; Herwitz, S. Human Challenges in the Maintenance of Unmanned Aircraft Systems; Interim Report to
FAA and NASA: NASA, Moffett Field, CA, USA, May 2006.
3. Overview of Military Drones Used By the UK Armed Forces; House of Commons Library: London, UK, 2015.
4. Bhamidipati, K.K.; Uhlig, D.; Neogi, N. Engineering Safety and Reliability into UAV Systems: Mitigating the
Ground Impact Hazard; University of Illinois, Urbana-Champaign: Urbana, IL, USA, August 2007; Volume
61822, doi:10.2514/6.2007-6510.
5. Austin, R. Unmanned Aircraft Systems; Wiley: Hoboken, NJ, USA, May 2010; ISBN 978-0-470-05819-0.
6. US Department of Defence. Electronic Reliability Design Handbook; Technical Report, MIL-HDBK-338B; US
Department of Defence: Washington, DC, USA, 1998.
7. Miller, J.A.; Minear, P.D.; Niessner, A.F.; DeLullo, A.M.; Geiger, B.R.; Long, L.N.; Horn, J.F. Intelligent
Unmanned Air Vehicle Flight Systems. In Proceedings of the AIAA 2005-7081 Infotech@Aerospace
Conference, Arlington, VA, USA, 35 September 2005.
8. Schmidt, J.; Parker, R. Development of A UAV Mishap Human Factors Database. In Proceedings of the
Unmanned Systems 1995 Proceedings, Washington, DC, USA, 1012 July 1995.
9. Ballenger, K. Unmanned Aircraft SystemsGeneral Overview. In Proceedings of the Presented to
American Institute of Aeronautics and Astronautics, San Diego, CA, USA, 2628 March 2013.
10. Paggi, R.; Mariotti, G.L.; Paggi, A.; Leccese, F. Optimization of Availability Operation via simulated
Prognostics. In Proceedings of the Metrology for Aerospace. In Proceedings of the 2nd IEEE International
Workshop, Benevento, Italy, 45 June 2015; pp. 4448, ISBN 978-1-4799-7568-6,
11. Clough, B.T. Unmanned Aerial Vehicles: Autonomous Control Challenges, A Researcher’s Perspective. J.
Aerosp. Comput. Inf. Commun. 2005, 2, 327347, doi:10.2514/1.5588.
Sensors 2018, 18, x FOR PEER REVIEW 14 of 16
12. Department of Defense. MIL-HDBK-217/F2 Reliability Prediction of Electronic Equipment; Department of
Defense: Washington, DC, USA, 1995.
13. De Francesco, E.; De Francesco, R. The CoDeF structure: A way to evaluate Ai including failures caused by
multiple minor degradations. In Proceedings of the 2nd IEEE International Workshop Metrology for
Aerospace, Benevento, Italy, 35 June 2015.
14. Weibel, R.; Hansman, R.J. Safety Considerations for Operation of Different Classes of UAVs in the NAS. In
Proceedings of the AIAA 3rd “Unmanned Unlimited” Technical Conference, Workshop and Exhibit,
Infotech@Aerospace Conferences, Chicago, IL, USA, 2023 September 2004; doi:10.2514/6.2004-6421.
15. Peroni, M.; Dolce, F.; Kingston, J.; Palla, C.; Fanfani, A.; Leccese, F. Reliability study for LEO satellites to
assist the selection of end of life disposal methods. In Proceedings of the 3rd IEEE International Workshop
on Metrology for Aerospace, MetroAeroSpace 2016Proceedings, Florence, Italy, 2123 June 2016; pp. 141
16. Kabir, A.; Bailey, C.; Lu, H.; Stoyanov, S. A review of data-driven prognostics in power electronics. In
Proceedings of the 35th International Spring Seminar on Electronics Technology, Bad Aussee, Austria, 9
13 May 2012; pp. 189192, doi:10.1109/ISSE.2012.6273136.
17. Heywood, J.B.; Sher, E. The Two-Stroke Cycle Engine: Its Development, Operation, and Design; Society of
Automotive Engineers, Inc.: Warrendale, PA, USA, 1999.
18. USAF Judge Advocate General’s Corps. USAF Accident Investigation Board Reports. Available online: (accessed on 01/07/2008).
19. US Department of Defense. Reliability Prediction of Electronic Components; Technical Report, MIL-HDBK-
217/F2; Department of Defense: Washington, DC, USA, 1991.
20. Smith, G.; Schroeder, J.B.; Navarro, S.; Haldeman, D. Development of a prognostics and health
management capability for the Joint Strike Fighter. In Proceedings of the 1997 IEEE Autotestcon
Proceedings AUTOTESTCON ‘97, IEEE Systems Readiness Technology Conference. Systems Readiness
Supporting Global Needs and Awareness in the 21st Century, Anaheim, CA, USA, 2225 September 1997;
pp. 676682, doi:10.1109/AUTEST.1997.643994.
21. Murtha, J.F. An Evidence Theoretical Approach to Design of Reliable Low-Cost UAV’s. Master’s Thesis,
Virginia Polytechnic Institute and State University, Blacksburg, VA, USA, 2009.
22. Liu, H.; Yu, J.; Zhang, P.; Li, S. A review on fault prognostics in integrated health management. In
Proceedings of the 2009 9th International Conference on Electronic Measurement & Instruments, Beijing,
China, 1619 August 2009; pp. 267270, doi:10.1109/ICEMI.2009.5274082.
23. Beard, R.W.; McLain, T.W. Small Unmanned AircraftTheory and Practice; Princeton University Press:
Princeton, NJ, USA, 2012.
24. Ng, Y.; Tomblin, J.; Freisthler, M. NCAMP Standard Operating Procedures. Available online:
20Operating%20Procedure%20March%2011% (accessed on 1 May 2012).
25. Paggi, R.; Mariotti, G.L.; Paggi, A.; Calogero, A.; Leccese, F. Prognostics via physics-based probabilistic
simulation approaches. In Proceedings of the 3rd IEEE International Workshop on Metrology for
Aerospace, MetroAeroSpace 2016, Florence, Italy, 2223 June 2016; pp. 130135.
26. De Oliveira Martins Franco, B.J.; Sandoval Góes, L.C. Failure Analysis Methods in Unmanned Aerial
Vehicle (UAV) Applications. In Proceedings of the Proceedings of COBEM 2007 19th International
Congress of Mechanical Engineering, Brasília, Brazil, 59 November 2007.
27. Schirripa Spagnolo, G.; Papalillo, D.; Leccese, F. Forensic Metrology: Uncertainty of Measurements in
Forensic Analysis. In Proceedings of the 20th IMEKO TC-4 International Symposium Measurement of
Electrical Quantities, Benevento, Italy, 1517 September 2014.
28. De Francesco, E.; De Francesco, R.; Leccese, F.; Cagnetti, M. Risk analysis in aviation: The forensic point of
view. In Proceedings of the 20th IMEKO TC4 Symposium on Measurements of Electrical Quantities,
Research on Electrical and Electronic Measurement for the Economic Upturn, together with 18th TC4
International Workshop on ADC and DCA Modeling and Testing, Benevento, Italy, 1517 September 2014,
pp. 563568.
29. Caciotta, M.; Cerqua, V.; Leccese, F.; Giarnetti, S.; DeFrancesco, E.; De Francesco, E.; Scaldarella, N. A First
Study on Prognostic System for Electric Engines Based on Envelope Analysis. In Proceedings of the IEEE
International Workshop on Metrology for Aerospace, Benevento, Italy, 2930 May 2014.
Sensors 2018, 18, x FOR PEER REVIEW 15 of 16
30. Blom, J.D. Unmanned Aerial System a Historical Perspective; Combat Studies Institute Press: Fort Leavenwort,
KS, USA, 2010. Available online:
(accessed on 5 April 2018).
31. ASD/AIA. S3000L International Procedure Specification for Logistic Support Analysis LSA. Issue 1.1. 24
July 2014. Available online: (accessed on 9 May 2016).
32. Kakaes, K.; Greenwood, F.; Lippincott, M.; Dosemagen, S.; Meier, P.; Wich, S. Drones and Aerial
Observation: New Technologies for Property Rights, Human Rights, and Global Development. New
America. 2015. Available online:
(accessed on 5 May 2016).
33. De Angelis, G.; Dati, E.; Bernabei, M.; Leccese, F. Development on aerospace composite structures
investigation using thermography and shearography in comparison to traditional NDT methods. In
Proceedings of the 2nd IEEE International Workshop on Metrology for Aerospace, MetroAeroSpace 2015,
Benevento, Italy, 45 June 2015; pp. 4955. doi:10.1109/MetroAeroSpace.2015.7180625.
34. Draper, M.; Calhoun, G.; Ruff, H.; Williamson, D.; Barry, T. Manual versus Speech Input for the Unmanned
Aerial Vehicle Control Station Operations. In Proceedings of the 47th Annual Meeting Human Factors and
Ergonomics Society, Denver, CO, USA, 1317 October 2003.
35. Tobon-Mejia, D.A.; Medjaher, K.; Zerhouni, N.; Tripot, G. A Data-Driven Failure Prognostics Method Based
on Mixture of Gaussians Hidden Markov Models. IEEE Trans. Reliab. 2012, 61, 491503,
36. Reimann, S.; Amos, J.; Bergquist, E.; Cole, J.; Phillips, J.; Shuster, S. UAV for Reliability Build; Technical
Report; Department of Aerospace Engineering and Mechanics, University of Minnesota, Minneapolis, MN,
USA, May 2014.
37. Bianchi, S.; Paggi, R.; Mariotti, G.L.; Leccese, F. Why and When Must the Preventive Maintenance be
Performed? In Proceedings of the IEEE International Workshop on Metrology for Aerospace, Benevento,
Italy, 2930 May 2014; pp. 221226.
38. USA Department of Defense. MIL-1388-2B Logistics Support Analysis Record. Available online: (accessed on 5 May 2016).
39. Wichita State University; SAE International. Polymer Matrix Composites: Guidelines for Characterization of
Structural Materials; ASTM International: West Conshohocken, PA, USA, 2002; Volume 1, ISBN 978-0-7680-
40. Saha, B.; Goebel, K.; Christophersen, J. Comparison of prognostic algorithms for estimating remaining
useful life of batteries. Trans. Inst. Meas. Control SAGE J. 2009, 31, 293308. Available online: (accessed on 5 May 2016).
41. Wood, S. Autonomous Underwater Gliders. In Underwater Vehicles; Inzartsev, A.V., Ed.; In-Tech: Vienna,
Austria, 2009; Chapter 26, pp. 499524.
42. Kayton, M.; Fried, W.R. Avionics Navigation Systems; Wiley: Hoboken, NJ, USA, 1997,
43. Wang, T.; Yu, J.; Siegel, D.; Lee, J. A similarity-based prognostics approach for Remaining Useful Life
estimation of engineered systems. In Proceedings of the International Conference on Prognostics and
Health Management, Denver, CO, USA, 69 October 2008; pp. 16, doi:10.1109/PHM.2008.4711421.
44. Schaefer, R. Unmanned Aerial Vehicle Reliability Study; Office of the Secretary of Defense: Washington, DC,
USA, February 2003.
45. Zio, E.; Di Maio, F. A data-driven fuzzy approach for predicting the remaining useful life in dynamic failure
scenarios of a nuclear system. Reliab. Eng. Syst. Saf. 2010, 95, 4957, doi:10.1016/j.ress.2009.08.001.
46. Petritoli, E.; Leccese, F.; Ciani, L. Reliability assessment of UAV systems. In Proceedings of the 2017 IEEE
International Workshop on Metrology for AeroSpace (MetroAeroSpace), Padua, Italy, 2123 June 2017; pp.
266270, doi:10.1109/MetroAeroSpace.2017.7999577.
47. Gnedenko, B.; Belyayev, Y.; Solovyev, A.D. Mathematical Methods of Reliability Theory; Barlow, R.E., Ed.;
Academic Press: New York, NY, USA, 1969.
48. Lloyd, D.; Lipow, M. Reliability: Management, Methods, and Mathematics. Am. Math. Mon. 1963, 70, 342.
49. Romig, H. Binomial Table; John Wiley and Sons, Inc.: New York, NY, USA, 1953.
50. Shooman, M. Probabilistic Reliability: An Engineering Approach; McGraw-Hill: New York, NY, USA, 1968.
51. Rao, C. Linear Statistical Inference and Its Applications; John Wiley and Sons, Inc.: New York, NY, USA, 1977.
Sensors 2018, 18, x FOR PEER REVIEW 16 of 16
52. Mann, N.; Schafer, R.; Sigpurwalla, N. Mathematical Methods for Statistical Analysis of Reliability and Life Data;
John Wiley and Sons, Inc.: New York, NY, USA, 1974.
53. Graver, J.G.; Liu, J.; Woolsey, C.; Leonard, N.E. Design and Analysis of an Underwater Vehicle for
Controlled Gliding. In Proceedings of the 1998 Conference on Information Sciences and Systems, Princeton,
NJ, USA, 1517 March 1998; pp. 801806.
54. Chen, Y.; Chen, T. Implementing Fault-Tolerance via Modular Redundancy with Comparison. IEEE Trans.
Reliab. 1990, 39, 217225, doi:10.1109/24.55885.
55. Feldstein, C.B.; Muzio, J.C. Development of a Fault Tolerant Flight Control System. In Proceedings of the
23rd Digital Avionics Systems Conference (IEEE Cat. No. 04CH37576), Salt Lake City, UT, USA, 28 October
2004; pp. 6361, doi:10.1109/DASC.2004.1390738.
56. Khan, R.; Williams, P.; Riseborough, P.; Rao, A.; Hill, R. Active Fault Tolerant Flight Control System
DesignA UAV Case Study. Available online: (accessed on 7 May
57. Ducard, G.J.J. Fault-tolerant Flight Control and Guidance Systems. In Practical Methods for Small Unmanned
Aerial Vehicles; Springer: Berlin, Germany, 2009; Volume XXII, p. 264.
58. Kobayashi, Y.; Takahashi, M. Design of Intelligent Fault-Tolerant Flight Control System for Unmanned
Aerial Vehicles. Keio University. Japan. Nihon Kikai Gakkai Ronbunshu C Hen/Trans. Jpn. Soc Mech Eng. Part
C 2009, 75, 23012310.
59. Zhang, X.; Li, H.; Yuan, D. Dual Redundant Flight Control System Design for Microminiature. UAV. In
Proceedings of the 2nd International Conference on Electrical, Computer Engineering and Electronics,
Jinan, China, 2931 May 2015, doi:10.2991/icecee-15.2015.153.
60. Department of the Army. Reliability/Availability of Electrical and Mechanical Systems for Command,
Control, Communications, Computer, Intelligence, Surveillance and Reconnaissance (C4ISR) Facilities. TM
5-698-1. 2003. Available online:
(accessed on 1 March 2018).
61. Poole, J. A Fast Reliability Analysis for an Unmanned Aerial Vehicle Performing a Phased Mission. Ph.D.
Thesis, Loughborough University, Loughborough, UK, 2011. Available online: (accessed on 1 March 2018).
62. Zhu, Q. Maintenance Optimization for Multi-Component Systems under Condition Monitoring.
Technische Universiteit Eindhoven, Eindhoven. 2015. Available online: (accessed on 1 March 2018).
63. Zhu, Q.; Peng, H.; Houtum, van Houtum, G.J. A condition-based maintenance policy for multicomponent
systems with a high maintenance setup cost. In OR Spectrum; Springer: Berlin, Germany, 2015; Volume 37,
pp. 10071035, doi:10.1007/s00291-015-0405-z.
64. DSIAC. Reliability of UAVs and Drones. Available online:
/spring-2017-volume-4-number-2/reliability-uavs-and-drones (accessed on 1 March 2018).
© 2018 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 (
... Certainly, there doesn't exist a unified method for handling all system problems, however, the proposed solution based on the FTC technique is appealing as it is dedicated to the system parts that are more likely to fail than others. According to [7], about 40% of UAVs failures are due to power plants faults and around 15% are resulting from the navigation system malfunction, which urges this work to investigate the problem of UAVs sensors and actuators faults aiming to propose an effective solution in a form of an active FTC algorithm that increases UAVs security and sustainability. ...
... In that figure, one can find the fault free, the system response without FTC, and its trajectory with the FTC law in green, blue, and red continuous lines, respectively. The Keeping in mind that the "o", "+" symbols represent the poles nominal and faulty positions, respectively, one can notice that the time response is degraded not only because of the shift that happened to the pole(3) in red which results in slower response, but also because of the conversion of the poles (7,8) from pure real poles to include an imaginary component which degrades the system damping. ...
Recently, autonomous systems are getting increasingly popular and are widely deployed in several applications in our daily life. That's why a great concern has been dedicated to the problem of autonomous systems fault-tolerant control (FTC). Evidently, the UAVs are among the systems that are in need of such FTC algorithms because any system malfunction can cause severe damage not just for the vehicle itself but for the surrounding environment as well. So this work is investigating the problem of designing an FTC algorithm for a quadrotor aiming to be a worthy contribution to the evolution of UAVs safety and reliability. Such a problem is tackled through some fundamental steps beginning with establishing a trustful model for the system representing the physical dynamics accurately. So Newton-Euler formulation is used for modeling the quadrotor resulting in a mathematical model that describes the relationship between the applied forces and the system states. After that the nonlinear model is linearized around the hovering point to simplify the control law design. A precise model could be constructed in an LPV framework where the nonlinear terms are considered as linearly time-varying within the given parameter limits. The deduced model is then used to build a controller that stabilizes the quadrotor and guarantees adequate trajectory tracking. So different types of control law are presented and analyzed some of them are linear controllers like PID provided with loop shaping technique. Other types of controllers presented are LQG to handle the system whose measurements are affected by Gaussian white noise and robust LPV control based on the H_inf technique to overcome unknown exogenous disturbances and measurement noise. In order to provide the quadrotor with an efficient FTC scheme, first, a fault detection and diagnosis (FDD) unit is proposed to identify the type, amount, and location of the existent fault. The FDD unit contains a model-based observer that generates some residual signals indicating the fault occurrence. According to the observer design, it may give just fault detection with a bank of observers for fault isolation or it can perform fault detection, estimation, and identification simultaneously. So an observer is designed based on H_/ H_inf technique aiming at maximizing the fault to residual sensitivity by using the H_ index properties, and minimizing the H_inf norm for worst-case exogenous signals attenuation. Afterward, a new approach is proposed for observer design based on an auxiliary output containing the system output and its successive time derivatives. This approach is used for both actuators and sensors fault diagnosis including fault detection, estimation, and isolation. It is illustrated that under some structural conditions, the faults can be estimated exactly while the perturbations are completely decoupled from the residual signals. However, if exact convergence is not ensured, some relaxed conditions are provided to maintain asymptotic fault estimation. Finally, the worst-case where the perturbations cannot be decoupled is presented and handled using H_/H_inf approach which is further enhanced utilizing the auxiliary output. Upon the obtained results from the actuator FDD unit, an active fault-tolerant control law is designed. After fault evaluation, the FDD gives a decision for the controller reconfiguration unit whether the actuator damage can be contained or not. For the first case, a control law is proposed aiming at fault compensation and precise trajectory tracking in the presence of system malfunction. For the latter case, a fail-safe mode is used to ensure that the quadrotor can land safely without crashing or causing harm to the surrounding environment.
... Another important property of UAVs is reliability, which allows determining the conditions of UAVs' functioning performance at the specified level [49]. The reliability analysis or reliability engineering is a knowledge domain, which includes many approaches, methods, and algorithms for reliability evaluation and risk assessment of complex systems. ...
Full-text available
UAVs have a great potential of application for monitoring, search, detection, communication, delivery and transportation of cargo in various sectors of economy. In spite of this, the existing software and hardware, as well as legal limitations, prevent the wide application of UAVs. There is intensive research related to automation and optimization of missions of one or more UAVs in various application areas. However, the execution of missions of both individual vehicles and their homogeneous or heterogeneous groups depends on the reliability issues of these technical devices, UAVs fleets, and control systems. In this paper, we present models for assessing fleet reliability of UAVs that are managed centralized or decentralized. The method is based on the representation of the fleet as a Binary-State System. The following topologies are considered: (a) a homogenous irredundant drone fleet, (b) a homogenous hot stable redundant drone fleet, (c) a heterogeneous irredundant drone fleet, and (d) a heterogeneous hot stable redundant drone fleet. For the listed topologies, reliability estimates were obtained as a function of the number of primary and redundant UAVs.KeywordsUAVsReliabilityBinary-State SystemImportance analysisStructure functionAvailability
... If we assume one step in the simulation corresponds to 10 seconds in the real world, the hourly failure rate of a patrolling system with two agents is approximately 0.34, which is much higher than the failure rate of real-world autonomous vehicles such as commercial UAVs [19]. As such, the effect of an agent failure in the real world will be less than the effect in these scenarios. ...
Full-text available
Autonomous vehicles are suited for continuous area patrolling problems. However, finding an optimal patrolling strategy can be challenging for many reasons. Firstly, patrolling environments are often complex and can include unknown and evolving environmental factors. Secondly, autonomous vehicles can have failures or hardware constraints such as limited battery lives. Importantly, patrolling large areas often requires multiple agents that need to collectively coordinate their actions. In this work, we consider these limitations and propose an approach based on a distributed, model-free deep reinforcement learning based multi-agent patrolling strategy. In this approach, agents make decisions locally based on their own environmental observations and on shared information. In addition, agents are trained to automatically recharge themselves when required to support continuous collective patrolling. A homogeneous multi-agent architecture is proposed, where all patrolling agents have an identical policy. This architecture provides a robust patrolling system that can tolerate agent failures and allow supplementary agents to be added to replace failed agents or to increase the overall patrol performance. This performance is validated through experiments from multiple perspectives, including the overall patrol performance, the efficiency of the battery recharging strategy, the overall robustness of the system, and the agents' ability to adapt to environment dynamics.
Full-text available
The article provides a brief overview of the history, types and functions of UAVs, its current and future areas of application, as well as a number of advantages and disadvantages.
One of the challenges of Unmanned Aircraft System (UAS) operations is to operate an unmanned aircraft with minimal risk to people on the ground. The purpose of this study is to define and measure such risks as population risk, by incorporating spatiotemporal changes in population density. Unlike previous studies, we use high-resolution de facto population data instead of residential population data to reflect the spatiotemporal characteristics of population distribution. Furthermore, we analyze the impact of mitigation measures based on population risk in the context of airspace management. We set a restricted airspace by using population risk and an acceptable level of safety. Scenario analysis of the study area in Seoul, South Korea provides a richer set of findings regarding spatiotemporal differences in restricted airspace. During the daytime, there are many restricted airspaces around commercial areas, but few around residential areas. Additionally, we observe the difference between restricting airspace based on population risk derived from the residential population and from the de facto population. These findings confirm the importance of accurately considering population density when assessing and mitigating the population risk associated with UAS operations. Sensitivity analysis also reveals the need to precisely estimate population density when estimating population risk with combinations of multiple parameter values. The proposed approach captures spatiotemporal characteristics of population distribution when assessing the population risk associated with UAS.
Full-text available
Fault tolerance is achieved through multiply redundant hardware systems in large civil aircraft. This means of achieving fault tolerance is infeasible for small compact unmanned aerial vehicles. In this paper we apply a fault tolerant control system which exploits analytical redundancy rather than hardware redundancy to an actual UAV model currently in operation via model-in-the-loop simulation. The fault tolerant control system comprises a nonlinear model predictive controller integrated with an unscented Kalman filter for fault detection and identification. The results show that our fault tolerant control system design is able to identify engine failure within seconds of fault occurrence and distribute control authority to the healthy actuators to maintain safe flight.
Conference Paper
Full-text available
Following a business as usual scenario, some Low Earth Orbit (LEO) regions could be unusable for many decades because of the space debris growth. In order to reduce that trend, the current probability of success of the chosen End of Mission (EOM) disposal method shall ensure a target value of 90% [1]. Understanding reliability of satellites and their subsystems for different spacecraft classes allows determining which disposal solution could better fit with a particular space mission. However, spacecraft are quite often different from each other, so a statistical approach is required. An in depth study has been performed on 1086 spacecraft launched between January 2000 and December 2014 using data from the SpaceTrak™ database. Spacecrafts have been separated by mass and by the presence/absence of the propulsion subsystem. The non-parametric Kaplan-Meier survival analysis has been used because the dataset presented censored events, namely the observed variable value is partially known. Empirical reliabilities obtained have been fitted using the Weibull distribution. Because each disposal method needs a combination of subsystems in order to operate, the reliabilities of the different subsystems have been combined by means of the System Reliability Theory. General spacecraft reliability was found to be about 92% after 4 years. The presence of the propulsion subsystem results in a better reliability trend. Furthermore, the propulsion presence/absence classification being equal, the heavier the mass the worse the reliability. Disposal solutions that use communication and power subsystems can count on reliabilities above 90% up to 7 years, whereas those ones that need also the attitude control can rely on only an 85% reliability after 4 years. A trade-off was performed and it showed that the film aerobrake and the propulsive D-Orbit decommissioning device can be key resources as disposal methods for future missions. The results presented could be useful to the space industry, to better address its efforts in improving spacecraft reliability and to design more reliable EOM disposal methods in order to reduce space debris growth.
This book addresses the two-stroke cycle internal combustion engine, used in compact, lightweight form in everything from motorcycles to chainsaws to outboard motors, and in large sizes for marine propulsion and power generation. It first provides an overview of the principles, characteristics, applications, and history of the two-stroke cycle engine, followed by descriptions and evaluations of various types of models that have been developed to predict aspects of two-stroke engine operation.
Some Applications of UASWhat are UAS?Why Unmanned Aircraft?The Systemic Basis of UASSystem CompositionReferences
This research aims at proposing an intelligent flight control system which makes unmanned aerial vehicles tolerant of actuator faults. The proposed system can detect, identify and accommodate the faults automatically. To make the unmanned system downsizing and low-cost, this study focuses on keeping the redundancy by software approach. The purpose of this study is to establish the systematic learning-based design method of control system with the evaluation function based on control purpose. The proposed technique adds to normal flight control system (Navigation, Guidance, and Control) with detection, identification, and accommodation mechanisms. Each mechanism consists of neural network. In this study, the availability of the proposed system is verified by six-degree-of-freedom nonlinear simulation. In the simulation, it was assumed that an unmanned aerial vehicle is in steady flight and the elevon fault (lock-in-place) is happened. Under various conditions, the proposed technique can evaluate the flight condition and decouple the broken actuator and generate a new flight path. Then, it can achieve the stable flight in spite of the actuator fault.
Conference Paper
Recent developments in the Logistic Engineering field are expanding the range of analysis for Operational Availability (Ao) from purely statistical models to include prognostic models. The prognostic method allows to obtain data on otherwise costly to test failure mechanics, while also providing information on the component degradation physics and their performance alteration. Failure in a system is commonly considered to be dependent on the detected malfunctioning of one or more items, as per the standard FMECA approach. A system failure though may also be caused by the concurrent degradation of multiple items performances which brings the system performance below a critical level. The performance degradation of those items is insufficient to trigger an alarm, which leads to cases of system failures with all components nominally in a working state. By using the degradation data obtained by the new prognostic models this paper introduces an analysis (named CoDeF) alternative to the RBD (Reliability Block Diagram). This approach is potentially able to take into consideration this latter type of system failure in the determination of the system Ai.