A Method to Analysis Maintenance Support

Capability of Aircraft Based on QFD

WEN Jing-Qian LI Qing

School of Mechanical Engineering & Automation School of Mechanical Engineering & Automation
Beihang University Beihang University
Beijing, China Beijing, China
[email protected] [email protected]

Abstract—It is a challenge for traditional index systems and find the essential reason to restrict the maintenance support
evaluation methods to reflect the relationship of maintenance capability.
support weakness and aircraft availability directly. The objective
of this research is to analysis the relevancy of support capability The method of QFD (Quality Function Deployment) can
and factors from the view of aircraft manufacture. The dig the user needs well [6,7], and transform the demand into
evaluation index system is established from two aspects, form corresponding management and control indexes effectively.
support capability aspect the indexes are composed of the Based on the method of QFD, this paper sets up the evaluation
reliability, maintenance and repair indexes, from maintenance and analysis method of aircraft maintenance support capability,
factor aspect the indexes are include personal, spare part, and divides indexes into assessing outcome index and
resource and management indexes. A new method to analysis reasonable management index to analyze the weakness of
maintenance support capability is put forward by combining maintenance support system. This provides an effective
MBO(management by objectives) and QFD(Quality Function analysis method in raising maintenance support level and
Deployment), and the evaluation index values and weights are improving product design.
calculated based on triangular fuzzy numbers. At last, an
example of this proposed methodology for certain type aircraft is
given to demonstrate its usefulness to improve maintenance and II. ANALYSIS METHOD BASED ON QFD
support capability analysis. Maintenance support capacity of equipment in the
condition of use and mission is not easy to measure. Most of
Keywords-QFD; capability evaluation; index system; triangular the equipment assessment studies use the allocation of
fuzzy numbers; maintenance support; maintenance support resources, maintenance process,
maintenance management and other external conditions as
I. INTRODUCTION indexes to assess indirectly. But the level of maintenance
Aircraft maintenance support is capability that maintaining support system is only the sufficient condition rather than the
and restoring the technical state of aircraft, improving aviation necessary condition for the successful completion of tasks.
equipment performance and training to support combat and Zhang Tao[3] sets up the indexes from operational units,
ensure perform other tasks successfully. To improve the equipment and maintenance support system, and gives the
maintenance support capability and achieve the design mathematical derivation for the mission success rate. It
usability of the aircraft by overcoming the weakness in provides theoretical analysis method for maintenance support
maintenance system, evaluation of maintenance support capability in design phase.
capability should not only contain the maintenance support For the analysis requirement of maintenance support
resources, but also pay attention to analyze the effect of capability in equipment use phases, the analysis of
maintenance support management. maintenance support capability should make support system
In the research of equipment maintenance support index conditions as the cause, indexes as the support result.
model, Yang Yi[1] builds the wartime index system of air force Quantitative assessment of support resources and management
communication equipment based on simulation method; Zuo level can find the weaknesses and deficiencies, so it can
Dehua[2] sets up the ship index system by its performance; provide the reference for managers. However, the evaluation
Zhang Tao[3] builds the comprehensive index system analysis of maintenance support capability in use phases has
according to support object and system . On the methods of difficulties as follows:
evaluation, Mu Fuling[4] gives the method of variable weight, • Hidden. Product requirement for customers are usually
Wang Yiyuan[5] evaluates the efficiency with rough set. The implied in the evaluation of maintenance support
research above mainly from the view of user evaluates static indexes.
ability and analyze effectiveness of maintenance support, but
lacks the cause analysis from the view of factory, so it’s hard to • Indirect. Maintenance resources and management
indexes can’t reflect the user needs directly.

978-1-4244-4905-7/09/$25.00©2009 IEEE
621
 • Fuzziness. The fuzzy semantic approach is usually
L11 A0, Availability

Maintenance support capability index

described the evaluation of qualitative indexes and
relative importance of different indexes. L12 RR, Readiness rate
For the problems above, according to the requirement of L13 MSR, Mission success rate
target management, a method of evaluation and analysis of
maintenance support capability based on QFD is put forward. L14 SGR, Sortie generation rate
That is meeting the demand of equipment maintenance support L21 FDR, Fault detection rate
in use phase as the target and comparing the design index with
the index based on index weight calculation. According to L22 FRR, Field repair rate
HOQ(House of Quality), the relationships between indexes of L31 MTBF, Mean Time Between Failure
equipment maintenance support capability and conditional
index satisfied maintenance support system resources will be L32 MTBR, Mean Time Between Removal
determined, the key factors and the weakness to affect L33 MTTR, Mean Time To Repair
equipment performance will be analyzed. The structure of
L34 TAT, Turnaround Time
HOQ in analysis of maintenance support capability is as
follows.
Figure 2. Aircraft maintenance support capability index

P1 Meet rate

Personal
P2 Specialized level

Maintenance support factor index

P3 Employable rate
S1 Meet rate

Spare
S2 Utilization
S3 Type march rate
S4 Quantity march rate
R1 Meet rate
Resource

Figure 1. HOQ for maintenance support capability analysis

R2 Utilization
① Equipment support index: support index is determined
by reliability, maintenance and support. R3 Software level
② Support factor index: index of support system, which M1 Service timely rate
Management

contains personnel, spare parts, process and M2 Average delayed time

management etc.
M3 Average response time
③ Relationship matrix: it reflects the index correlation
and coverage degree between equipment support M4 Average processing time
index and support factor index. Figure 3. Aircraft maintenance support capability index
④ Autocorrelation matrix: it reflects the degree of
mutual influence between support factor indexes, and The weight of equipment support index in accordance with
requires avoiding strong correlation. the target management method compares with the support
design demand. Index weight is identified by the differential,
⑤ Equipment support competitive analysis: analyze the support factor index is got by calculating the relationship
differences of equipment support among different matrix.
types and different users.
⑥ Support capability competitive analysis: analyze the III. THE INDEX WEIGHT CALCULATION METHOD BASED
differences of maintenance support capability among ON TRIANGLE FUZZY NUMBERS
different maintenance support units. Aircraft maintenance support indexes contain the form of
ratio, numbers, and descriptive semantic value and so on.
Building HOQ of maintenance support capacity analysis,
Considering the expression of index and comparative
equipment maintenance support index is conformed by users
calculation, the paper uses triangle fuzzy numbers to describe.
and supporters. Support factor index is determined by
For a n-dimensional fuzzy evaluation, we can establish a
supporters in according with maintenance support capacity
triangle fuzzy set:
analysis and management requirement, and the coverage
degree of support factor index in relationship matrix to support I = (i1,…,in ) (1)
index is given by experts and professionals. The common
aircraft maintenance support capability index is as follows:

622
im can be expressed as: [Step1] A survey is conducted to k customers for n
support ability indexes. The index description set given by the
m −1 m m +1 actual equipment support ability indexes is:
( , , )
X = { X 1, X 2 , X k }
(2)
n n n (4)
excluding the fuzzy value:
The information of prejudice and favoritism of evaluator
l + 2m + u used weighted method is:
e=
4 (3) P = { p1,…, pn } (5)
some corresponding relationships are shown in the table below:
pi = r1xi1 ⊕ r2 xi2 …⊕ rk xik , i = 1,…,n . (6)
common better good Very good excellent r is the weight parameter for the importance of the
evaluator:
Variables
k

rules ∑r =1
j =1
j (7)

[Step2] The designed equipment support index:

0.375 0.5 0.625 0.75 0.875 1 field
Figure 4. Distribution of the Triangle Fuzzy Numbers T = {t1,…, tn} (8)

Work out the gap between the survey and the target:
TABLE I. LINGUISTIC VARIABLES COMPARISON
D = {d1,…, dn } (9)
Comparison Variable Triangle Fuzzy Numbers
Equal Common (0.375, 0.5, 0.625) di = e( pi ) − e(ti ) (10)
Less Important Good (0.5, 0.625, 0.75)
di is the margin between ti and pi after taking off fuzzy.
Important Better (0.625, 0.75, 0.875) It sets up the index weight matrix of comparison precedence
More Important Best (0.75, 0.875, 1) relationship: Fn×n , where:
Very Important Perfect (0.875, 1, 1)
⎧0 di < d j
⎪
Give the fuzzy description of linguistic index on [0, 1], “an
fij = ⎨1 di > d j (11)
80% of satisfaction” can be described as (0.75, 0.80, 1) or ⎪0.5 d = d
(0.625, 0.80, 0.875). For the index which is not on [0, 1], ⎩ i j

should be transformed to [0, 1], and then described as the form

of fuzzy numbers. For example, the MTBF (Mean Time [Step3] Turn Fn×n into fuzzy consistent matrix:
Between Failures) of a spare is “about 730 h ”, the effective
Rn ×n = ( rij ) n ×n (12)
time range is[200 h ,1500 h ], and transform it on [0, 1] by the
method of trapezoid fuzzy function, then it can be described as
the expression of rij is:
(0.375, 0.408, 0.625).
Triangular fuzzy numbers[8,9] is not only able to quantify ri − rj
the semantic description, but also realize the same description rij = + 0.5 (13)
of different indexes, which provide a good data base for the 2n
compare between different indexes. On the base of triangular n
ri = ∑ fik , i = 1,
fuzzy numbers, this paper improves the traditional method of
weight determine in expert scoring. According to target ,n (14)
k =1
management method, this paper provides a method to fix on
evaluation index weight based on the gap of equipment support [Step4] Using SqrtMethod for Eigenvalue, after normalize
degree of the customer. It compares the designed support index get the user needs indexes weight set:
with the actual result, and gives the relative index weight by
difference. The analysis steps of evaluation index weight of the Wn = (w1,…, wn )T (15)
maintenance support capability based on triangular fuzzy
numbers is given as follows:

623
 n {50%-100%, 50%-100%, 50%-100%, 1-4 times/day, 50%-
wi = wi′ × (∑w′j )−1 (16) 100%, 50%-100%, 1h-10h, 0.5h-2h, 1h-4h, 0.5h-2h};
j =1 The actual value of the maintenance indexes of the two
n
users as follows:
wi′ = (∏rij )
1
n
(17) Xa ={74%, 60%, 92%, 2.1times/day, 76%, 69%, 5.8h, 1.1h,
j =1 2.1h, 1.1h};
[Step5] By analyzing the association, get the relation Xb ={82%, 75%, 83%, 2.5times/day, 63%, 56%, 3.6h, 1.4h,
a b 3.1h, 1.2h};
matrixes of two periods house of quality: A n×m and A m×v , then:
For the two users, take 0.5 as the weigh coefficient, the
Factor indexes weight: weights of maintenance support capability indexes can be
calculated by the method as above:
Wma = (w1a ,…, wma )T (18)
W = {0.106, 0.150, 0.048, 0.138, 0.095, 0.117, 0.128,
Wma = Ana×m ×Wn (19) 0.073, 0.084, 0.061};
The first three indexes, which lower than the expected
According to these steps, the gap between designed support
value, are aircraft availability, Sortie generation rate and Mean
ability and the actual result can be transformed into the weight
Time Between Failure. By the HOQ, the weights of the
sets of support ability indexes and support factor indexes. The
maintenance factors indexes can be calculated:
bigger the gap is, the longer the distance to anticipation is. This
point should be focused on and improved.
Wa ={0.040, 0.126, 0.092, 0.109, 0.047, 0.095, 0.059,
0.046, 0.020, 0.033, 0.069, 0.092, 0.080};
IV. APPLICATION OF ANALYSIS METHOD
therefore, the personal ability, the part meet rate and variety
Base on an aircraft manufacture enterprise, using the match rate should be taken as the key items among the
method of QFD, the maintenance support capability of two maintenance factors indexes.
different users A and B is evaluated and analyzed, referring the
index system of Table Ⅱ, the relevant matrix value, given by
experts, is shown as follows: CONCLUSIONS
In this paper, the maintenance support capability analysis
method aimed at narrowing the gap between design and reality,
TABLE II. THE RELEVANT MATRIX OF THE CAPABILITY INDEXES AND
THE FACTORS INDEXES take the difference as reference to calculate the index weights,
and take the advantage of QFD to find out the weakness in
Maintenance support factor indexes maintenance support. The maintenance requirement in the
P1 P2 P3 S1 S2 S3 S4 R1 R2 R3 M1 M2 M3 M4 usage phase of aircraft changes continuously, the supporter can
take the advantage of the analysis method to synthetically
L11 0 0 0 7 3 7 5 0 0 0 9 7 7 7
Maintenance support capability indexes

improve maintenance support management capability by

L12 0 0 0 7 3 5 3 0 0 0 7 5 5 5 adjusting strategies timely.
L13 0 0 0 3 1 3 1 0 0 0 0 0 0 0 Because the maintenance factor index weights are
L14 0 0 0 3 1 3 1 0 0 0 3 1 3 1 calculated by HOQ, the relevant matrix has great effect on the
result, so how to correctly evaluate the relevant value in the
L21 3 9 7 0 0 0 0 7 3 5 0 0 0 0 matrix should be further study. Moreover, the impact of the
L22 5 9 7 0 0 0 0 7 3 5 0 0 0 0 relevance among indexes is not considered in this page which
is another research.
L31 0 5 3 0 0 0 0 0 0 0 0 0 0 0
L32 0 7 5 0 0 0 0 0 0 0 0 0 0 0 REFERENCES
L33 3 7 5 9 3 7 5 0 0 0 7 7 9 9 [1] YANG Yi, LIU Han, WU Chang, QI Sheng-li. Study of Effectiveness
Evaluation of Maintenance Support System Based on Fractal
L34 3 7 5 7 5 7 5 0 0 0 0 0 5 3
Conception[J]. Journal of System Simulation, 2006,18 (10):2925-2928
Remark: the relevant value is figured between 0 and 9, 0 means [2] Zuo Dehua, Li Junhua, Li Ming. Research on Effectiveness Evaluation
no relevance, 9 means strong. Method of Ship' s Maintenance Equipment[J]. Journal of Wuhan
University of Technology (Transportation Science &Engineering),
The expected target of maintenance index values is T: 2005,29 (1):34-36
[3] ZHANG Tao, ZHANG Jian-jun, GUO Bo. TAN Yue-jin. Study on
T ={90%, 85%, 90%, 1.5 times/day, 80%, 75%, 7.5h, 1h, Index Sytem for Maintenance Support Capability Assessment of
Equipment in Application Phase[J]. Journal of Academy of Armored
2h, 1h};
Force Engineering,2005,19 (1): 7-10
The corresponding value range is : [4] MUFu-ling, WU Chang, WU De-wei. Study on the Synthetic Method of
Variable Weight of Effectiveness Evaluation of Maintenance Support
System[J]. Systems Engineering and Electronics,2003,25 (6): 693-696

624
[5] Wang Liyuan, Zhang Hengxi, Xu Haojun. Multi-index Synthesize (2): 380-390
Evaluation Model Based on Rough Set Theory for Air Combat [8] Si Shubin, Sun Shudong, Han Guangchen, Wang Junqiang. A Weighted
Efficiency[J]. ACTA AERONAUTICA ET ASTRONAUTICA Analysis of an Integrated Logistics Support Based on Triangular Fuzzy
SINICA,2008,29 (4): 880-885 Number[J]. Journal of Northwestern Polytechnical University,2004,
[6] Herzwurm, G., Schockert, S., Pietsch, W.. QFD for customer-focused 22(6):689-694
requirements engineering[J]. Requirements Engineering Conference 8- [9] CHEN Xiao-hong, YANG Xi. Multiple attributive group decision
12 Sept. 2003:330-338 making method based on triangular fuzzy numbers [J]. Systems
[7] Sireli, Y., Kauffmann, P., Ozan, E.. Integration of Kano's Model Into Engineering and Electronics, 2008 30(2):278-282.
QFD for Multiple Product Design[J]. Engineering Management, 2007,59

625