Nof S.Y. Springer Handbook of Automation

Подождите немного. Документ загружается.

Automating Errors and Conﬂicts Prognostics and Prevention 30.5 Error Recovery and Conﬂict Resolution 515

work protocols) and conﬂict and error detection agents

(CEDAs, part of MAS) has been developed [30.91]to

detect errors and conﬂicts in different network topolo-

gies. The CEDM integrates CEDP, CEDAs, and four

error and conﬂict detection components (Fig.30.11).

A CEDA is deployed at each unit of a system to

(1) detect errors and conﬂicts by three components (de-

tection policy generation, error detection, and conﬂict

evaluation), which interact with and are supported by

error knowledge base, and (2) communicate with other

CEDAs to send and receive error and conﬂictannounce-

ments with the support of CEDP. The CEDM has been

applied to four different network topologies and the re-

sults show that the performance of CEDM is sometimes

counterintuitive, i. e., it performs better on networksthat

seem morecomplex. To be ableto detect both errors and

conﬂicts is desired when they exist in the same system.

Because errors are different from conﬂicts, the activi-

ties to detect them are often different and need to be

integrated.

30.4.3 Performance Measures

Performance measures are necessary for the evaluation

and comparison of various error and conﬂict prognos-

tics and prevention methods. Several measures have

already been deﬁned and developed in previous re-

search:

1. Detection latency: The time between the instant that

an error occurs and the instant that the error is de-

tected [30.10,91].

2. Error coverage: The percentage of detected errors

with respect to the total number of errors [30.10].

3. Cost: The overhead caused by including error de-

tection capability with respect to the system without

the capability [30.10].

4. Conﬂict severity: The severity of a conﬂict. It is the

sum of the severity caused by the conﬂict at each

involving unit [30.91].

5. Detectability: The ability of a detection method.

It is a function of detection accuracy, cost, and

time [30.92].

6. Preventability:The ratioof thenumber of errors pre-

vented divided by the total number of errors [30.80].

7. Reliability: The ratio of the number of errors pre-

vented divided by the number of errors identiﬁed or

predicted, or the ratio of the number of errors de-

tected divided by the total number of errors [30.40,

45,80].

Other performance measures, e.g., total damage and

cost–beneﬁt ratio, can be developed to compare differ-

ent methods. Appropriate performance measures help

determine how a speciﬁc method performs in different

situations and are often requiredwhen thereare multiple

methods available.

30.5 Error Recovery and Conﬂict Resolution

When an error or a conﬂict occurs and is detected,

identiﬁed, isolated or diagnosed, there are three possi-

ble consequences: (1) other errors or conﬂicts that are

caused by the error or conﬂict have occurred; (2) other

errors or conﬂicts that are caused by the error or con-

ﬂict will (probably) occur; (3) other errors or conﬂicts,

or the same error or conﬂict, will (probably) occur if

the current error or conﬂict is not recovered or resolved,

respectively. One of the objectives of error recovery

and conﬂict resolution is to avoid the third consequence

when an error or a conﬂict occurs. They are therefore

part of error and conﬂict prognostics and prevention.

There has been extensive research on automated er-

ror recovery and conﬂict resolution, which are often

domain speciﬁc. Many methods have been developed

and applied in various real-world applications in which

the main objective of error recovery and conﬂict res-

olution is to keep the production or service ﬂowing;

for instance, Fig. 30.12 shows a recovery tree for rheo-

stat pick-up and insertion, which is programmed for

automatic error recovery. Traditionally, error recovery

and conﬂict resolution are not considered as an ap-

proach to prevent errors and conﬂicts. In the next two

Sections, we describe two examples, error recovery in

robotics [30.93] and conﬂict resolution in collaborative

facility design [30.88,94], to illustrate how to perform

these two functions automatically.

30.5.1 Error Recovery

Error recovery cannot be avoided when using robots

because errors are an inherent characteristic of robotic

applications [30.95] that are often not fault tolerant.

Most error recovery applications implement prepro-

grammed nonintelligent corrective actions [30.95–98].

Due to the large number of possible errors and the

Part C 30.5

516 Part C Automation Design: Theory, Elements, and Methods

Move to next rheostat

Release rheostat

Move to next rheostat

Recalibrate robot

Return to start

Freeze; Call

operator

Return to start

Go to next feeder

Return to start

Freeze; Call

operator

Search for

insertion

Return to start

Freeze; Call

operator

Return to start

Recalibrate

robot

Set failure-

counter to 0

Return to start

Is feeder aligned?

Return to start

Discard

rheostat

Increment

failure-

counter

Counter <3?

Return to start

Start

Rheostat positioned

correctly?

Rheostat

available?

Rheostat

not caught?

Rheostat

inserted?

End

Fig. 30.12 Recovery tree for rheostat pick-up and insertion recovery. A branch may only be entered once; on success

branch downward; on failure branch to right if possible, otherwise branch left; when the end of a branch is reached,

unless otherwise speciﬁed return to last sensing position; “?” signiﬁes sensing position where sensors or variables are

evaluated (after [30.1])

inherent complexity of recovery actions, to automate

error recovery fully without human interventions is dif-

ﬁcult. The emerging trend in error recovery is to equip

Table 30.4 Multiapproach conﬂict resolution in collab-

orative design (Mcr) structure [30.88,94] (after [30.94],

courtesy Elsevier, 2008)



Part C 30.5

Automating Errors and Conﬂicts Prognostics and Prevention 30.5 Error Recovery and Conﬂict Resolution 517

Stage Strategy Steps to achieve conﬂict resolution Methodologies and tools

Mcr(1) Direct

negotiation

1. Agent prepares resolution proposaland sends to counterparts

2. Counterpart agents evaluate proposal. If they accept it, go to

step 5; otherwise go to step 3

3. Counterpart agents prepare a counteroffer and send it back

to originating agent

4. Agent evaluates the counteroffer. If accepted go to step 5;

otherwise go to Mcr(2)

5. End of the conﬂict resolution process

Heuristics;

Knowledge-based

interactions;

Multiagent systems

Mcr(2) Third-party

mediation

1. Third-party agent prepares resolution proposal and sends to

counterparts

2. Counterpart agents evaluate the proposal. If accepted, go to

step 5; otherwise go to step 3

3. Counterpart agents prepare a counteroffer and send it back

to the third-party agent

4. Third-party agent evaluates counteroffer. If accepted go to

step 5; otherwise go to Mcr(3)

5. End of the conﬂict resolution process

Heuristics;

Knowledge-based

interactions;

Multiagent systems;

PESUADER [30.99]

Mcr(3) Incorporation

additional

parties

1. Specialized agent prepares resolution proposal and sends to

counterparts

2. Counterpart agents evaluate the proposal. If accepted, go to

step 5; otherwise go to step 3

3. Counterpart agents prepare a counteroffer and send it back

to the specialized agent

4. Specialized agent evaluates counteroffer. If accepted go to

step 5; otherwise go to Mcr(4)

5. End of the conﬂict resolution process

Heuristics;

Knowledge-based

interactions;

Expert systems

Mcr(4) Persuasion

1. Third-party agent prepares persuasive arguments and sends

to counterparts

2. Counterpart agents evaluate the arguments

3. If the arguments are effective, go to step 4; otherwise go to

Mcr(5)

4. End of the conﬂict resolution process

PERSUADER [30.99];

Case-based reasoning

Mcr(5) Arbitration

1. If conﬂict management and analysis results in common pro-

posals (X), conﬂict resolution is achieved through management

and analysis

2. If conﬂict management and analysis results in mutually ex-

clusive proposals (Y), conﬂict resolution is achieved though

conﬂict confrontation

3. If conﬂict management and analysis results in no conﬂict

resolution proposals (Z), conﬂict resolution must be used

Graph model for

conﬂict resolution

(GMCR) [30.100]for

conﬂict management and

analysis;

Adaptive neural-fuzzy

inference system

(ANFIS) [30.101]for

conﬂict confrontation;

Dependency

analysis [30.102]and

product ﬂow analysis for

conﬂict resolution

Part C 30.5

518 Part C Automation Design: Theory, Elements, and Methods

Table 30.5 Summary of error and conﬂict prognostics and prevention theories, applications, and open challenges

Applications Assembly and Process monitoring Hardware testing Software testing

inspection

Methods/ Control theory; Analytical Data-driven Knowledge- Information theory; Model checking;

technologies

Knowledge base; based Heuristic search Bogor;

Computer/machine Cadena;

vision; Concurrent error

Robotics; detection (CED)

Feature extraction;

Pattern recognition

Functions

Detection × × × × × ×

Diagnostics × × × × ×

Identiﬁcation × × × × × ×

Isolation × × × × × ×

Error recovery ×

Conﬂict

resolution

Prognostics × × ×

Exception ×

handling

Errors/conﬂicts E E E E E E

Centralized/ C C C C C C

decentralized

Strengths Integration of error Accurate and Can process Does not Accurate Thorough

detection and reliable large amount require detailed and veriﬁcation

recovery of data system reliable with formal methods

information

Weaknesses Domain speciﬁc; Require Rely on the Results are Difﬁcult to derive State explosion;

Lack of general mathematical quantity, subjective and optimal algorithms Duplications needed

methods models that quality, and may not be to minimize cost; in CED;

are often timeliness reliable Time consuming Cannot deal with

not available of data for large systems incorrect or

incomplete

speciﬁcations

[30.3,20–34] [30.17,19,35] [30.36–45] [30.8,9,14,47–57]

systems with human intelligence so that they can cor-

rect errors through reasoning and high-level decision

making. An example of an intelligent error recovery

system is the neural-fuzzy system for error recovery

(NEFUSER) [30.93]).

The NEFUSER is both an intelligent system and

a design tool of fuzzy logic and neural-fuzzy models

for error detection and recovery. The NEFUSER has

been applied to a single robot working in an assem-

bly cell. The NEFUSER enables interactions among

the robot, the operator, and computer-supported ap-

plications. It interprets data and information collected

by the robot and provided by the operator, analyzes

data and information with fuzzy logic and/or neural-

fuzzy models, and makes appropriate error recovery

decisions. The NEFUSER has learning ability to im-

Part C 30.5

Automating Errors and Conﬂicts Prognostics and Prevention 30.5 Error Recovery and Conﬂict Resolution 519

Table 30.5 (cont.)

Applications Discrete event Collaborative design Production and service

system

Methods/ Petri net; Intended goal Facility Detection and Conﬂict and Fuzzy logic;

technologies

Finite-state structure (IGS); description prevention error detection Artiﬁcial

machine (FSM) Project evaluation language algorithms; model (CEDM); intelligence

and review (FDL); Reliability Active

technique (PERT); Mcr; theory; middleware

Petri net; CDMS Process

Conﬂict detection modeling;

and management Workﬂow

system (CDMS)

Functions

Detection × × × ×

Diagnostics

Identiﬁcation × × × ×

Isolation × × × ×

Error recovery ×

Conﬂict ×

resolution

Prognostics × × ×

Exception × ×

handling

Errors/conﬂicts E C C E E/C E

Centralized/ C/D C/D C/D C/D D C/D

decentralized

Strengths Formal method Modeling Integration of Reliable; Short detection Correct errors

applicable to of systems traditional human Easy to apply time through

various with agent-based conﬂict reso- reasoning and

systems technology lutions and high-level

computer-based decision making

learning

Weaknesses State explosion An agent The adaptability Limited to Needs further Needs further

for large may be modeled of the methods to sequential development development

systems; for multiple times other design production and and for various

System due to many activities has not service lines; validation applications

modeling is conﬂicts been validated Domain

complex and it is involved speciﬁc

time-consuming

[30.58–67] [30.66,82–88] [30.77,86,88,94] [30.68–80] [30.77,89–91] [30.93,95–98,103]

[30.104–114]

prove corrective actions and adapt to different errors.

The NEFUSER therefore increases the level of automa-

tion by decreasing the number of times that the robot

has to stop and the operator has to intervene due to

errors.

Figure 30.13 shows the interactions between the

robot, the operator, and computer-supported applica-

tions. The NEFUSER is the error recovery brain and

is programmed and run on MATLAB, which provides

a friendly windows-oriented fuzzy inference system

(FIS) that incorporates the graphical user interface tools

of the fuzzy logic toolbox [30.103]. The example in

Fig.30.13 includes a robot and an operator in an as-

sembly cell. In general, the NEFUSER design for error

Part C 30.5

520 Part C Automation Design: Theory, Elements, and Methods

NEFUSER

Request

help

Interact

with the

system

Recovery

strategy

Recovery

instructions

Operations and

recovery actions

Assist

Sensing

Sensor data

Sensor

information

Operator

Controller

Robot

Production

process

Fig. 30.13 Interactions with NEFUSER (after [30.93])

recovery includes three main tasks: (1) design the FIS,

(2) manage and evaluate information, and (3) train the

FIS with real data and information.

30.5.2 Conﬂict Resolution

There is a growing demand for knowledge-intensive

collaboration in distributed design [30.94, 113, 114].

Conﬂict detection has been studied extensively in col-

laborative design, as has conﬂict resolution, which

is often the next step after a conﬂict is detected.

There has been extensive research on conﬂict resolution

(e.g., [30.105–110]). Recently, a multiapproach method

to conﬂict resolution in collaborative design has been

introduced with the development of the facility descrip-

tion language–conﬂict resolution (FDL-CR) [30.88].

The critical role of computer-supported conﬂict resolu-

tion in distributed organizations has been discussed in

great detail [30.77, 104, 111, 112]. In addition, Ceroni

and Velasquez [30.86] have developed the conﬂict de-

tection and management system (CDMS)andtheir

work shows that both product complexity and number

of participating designers have a statistically signiﬁcant

effecton theratio ofconﬂicts resolved tothose detected,

but that only complexity had a statistically signiﬁcant

effect on design duration.

Based on the previous work, most recently, a new

method, Mcr (Table 30.4), has been developed to

automatically resolve conﬂict situations common in

collaborative facility design using computer-support

tools [30.88, 94]. The method uses both traditional

human conﬂict-resolution approaches that have been

used successfully by others and principles of con-

ﬂict prevention to improve design performance and

apply computer-based learning to improve usefulness.

A graph model for conﬂict resolution is used to facil-

itate conﬂict modeling and analysis. The performance

of the new method has been validated by implementing

its conﬂict resolution capabilities in the FDL,acom-

puter tool for collaborative facility design, and by

applying FDL-CR, to resolve typical conﬂict situations.

Table 30.4 describes the Mcr structure.

Table 30.5 summarizes error and conﬂict prognos-

tics and prevention methods and technologies in various

production and service applications.

30.6 Emerging Trends

30.6.1 Decentralized and Agent-Based

Error and Conﬂict Prognostics

and Prevention

Most error and conﬂict prognostics and prevention

methods developed so far are centralized approaches

(Table 30.5) in which a central control unit controls

data and information and executes some or all eight

functions to prevent errors and conﬂicts. The central-

ized approach often requires substantial time to execute

various functions and the central control unit often

possesses incomplete or incorrect data and informa-

tion [30.80]. These disadvantages become apparent

when a system has many units that need to be examined

for errors and conﬂicts.

To overcome the disadvantages of the centralized

approach, the decentralized approach that takes advan-

tage of the parallel activities of multiple agents has been

developed [30.16, 67, 79, 80, 91]. In the decentralized

approach, distributed agents detect, identify or isolate

Part C 30.6

Automating Errors and Conﬂicts Prognostics and Prevention 30.7 Conclusion 521

errors and conﬂicts at individual units of a system, and

communicate with each other to diagnose and prevent

errors and conﬂicts. The main challenge of the decen-

tralized approach is to develop robust protocols that

can ensure effective communications between agents.

Further research is needed to develop and improve de-

centralized approaches for implementation in various

applications.

30.6.2 Intelligent Error and Conﬂict

Prognostics and Prevention

Compared with humans, automation systems perform

better when they are used to prevent errors and

conﬂicts through the violation of speciﬁcations or vi-

olation in comparisons [30.13]. Humans, however, have

the ability to prevent errors and conﬂicts through

the violation of expectations, i.e., with tacit knowl-

edge and high-level decision making. To increase the

effectiveness degree of automation of error and con-

ﬂict prognostics and prevention, it is necessary to

equip automation systems with human intelligence

through appropriate modeling techniques such as fuzzy

logic, pattern recognition, and artiﬁcial neural net-

works. There has been some preliminary work to

incorporate high-level human intelligence in error de-

tection and recovery (e.g., [30.3, 93]) and conﬂict

resolution [30.88, 94]. Additional work is needed to

develop self-learning, self-improving artiﬁcial intelli-

gence systems for error and conﬂict prognostics and

prevention.

30.6.3 Graph and Network Theories

The performance of an error and conﬂict prognostics

and prevention method is signiﬁcantly inﬂuenced by

the number of units in a system and their relationship.

A system can be viewed as a graph or a network with

many nodes, each of which represents a unit in the sys-

tem. The relationship between units is represented by

the link between nodes. The study of network topolo-

gies has a long history stretching back at least to the

1730s. The classic model of a network, the random net-

work, was ﬁrst discussed in the early 1950s [30.115]

and was rediscovered and analyzed in a series of papers

published in the late 1950s and early 1960s [30.116–

118]. Most recently, several network models have been

discovered and extensively studied, for instance, the

small-world network (e.g., [30.119]), the scale-free

network (e.g., [30.120–123]), and the Bose–Einstein

condensation network [30.124]. Bioinspired network

models for collaborative control have recently been

studied by Nof [30.125] (see also Chap.75 for more

details).

Because the same prognostics and prevention

method may perform quite differently on networks with

different topologies and attributes, or with the same

network topology and attributes but with different pa-

rameters, it is imperative to study the performance of

prognostics and prevention methods with respect to dif-

ferent networks for the best match between methods

and networks. There is ample room for research, de-

velopment, and implementation of error and conﬂict

prognostics and prevention methods supported by graph

and network theories.

30.6.4 Financial Models

for Prognostics Economy

Most errors and conﬂicts must be detected, isolated,

identiﬁed, diagnosed or prevented. Certain errors and

conﬂicts, however, may be tolerable in certain systems,

i.e., fault-tolerant systems. Also, the cost of automating

some or alleight functionsof errorand conﬂictprognos-

tics and prevention may far exceed the damages caused

by certain errors and conﬂicts. In both situations, cost–

beneﬁt analyses can be used to determine if an error or

a conﬂict needs to be dealt with. In general, ﬁnancial

models are used to analyze the economy of prognostics

and prevention methods for speciﬁc errors and con-

ﬂicts, to help decide which of the eight functions will

be executed and how they will be executed, e.g., the fre-

quency. There has been limited research on how to use

ﬁnancial models to help justify the automation of er-

ror and conﬂict prognostics and prevention [30.92,126].

One of the challenges is how to appropriately evalu-

ate or assess the damage of errors and conﬂicts, e.g.,

short-term damage, long-term damage, and intangible

damage. Additional research is needed to address these

economical decisions.

30.7 Conclusion

In this Chapter we have discussed the eight func-

tions that automate error and conﬂict prognostics and

prevention and their applications in various produc-

tion and service areas. Prognostics and prevention

Part C 30.7

522 Part C Automation Design: Theory, Elements, and Methods

methods for errors and conﬂicts are developed based

on extensive theoretical advancements in many sci-

ence and engineering domains, and have been suc-

cessfully applied to various real-world problems. As

systems and networks become larger and more com-

plex, such as global enterprises and the Internet,

error and conﬂict prognostics and prevention be-

come more important and the focus is shifting from

passive response to active prognostics and preven-

tion.

References

30.1 S.Y. Nof, W.E. Wilhelm, H.-J. Warnecke: Industrial

Assembly (Chapman Hall, New York 1997)

30.2 L.S. Lopes, L.M. Camarinha-Matos: A machine

learning approach to error detection and recovery

in assembly, Proc. IEEE/RSJ Int. Conf. Intell. Robot.

Syst. 95, ’Human Robot Interaction and Coopera-

tive Robots’, Vol. 3 (1995) pp. 197–203

30.3 H. Najjari, S.J. Steiner: Integrated sensor-based

control system for a ﬂexible assembly, Mechatron-

ics 7(3), 231–262 (1997)

30.4 A. Steininger, C. Scherrer: On ﬁnding an optimal

combination of error detection mechanisms based

on results of fault injection experiments, Proc. 27th

Annu. Int. Symp. Fault-Toler. Comput., FTCS-27,

Digest of Papers (1997) pp. 238–247

30.5 K.A. Toguyeni, E. Craye, J.C. Gentina: Framework to

design a distributed diagnosis in FMS, Proc. IEEE

Int. Conf. Syst. Man. Cybern. 4, 2774–2779 (1996)

30.6 J.F. Kao: Optimal recovery strategies for manufac-

turing systems, Eur. J. Oper. Res. 80(2), 252–263

(1995)

30.7 M. Bruccoleri, Z.J. Pasek: Operational issues in

reconﬁgurable manufacturing systems: exception

handling, Proc. 5th Biannu. World Autom. Congr.

(2002)

30.8 T. Miceli, H.A. Sahraoui, R. Godin: A metric based

technique for design ﬂaws detection and correc-

tion, Proc. 14th IEEE Int. Conf. Autom. Softw. Eng.

(1999) pp. 307–310

30.9 C. Bolchini, W. Fornaciari, F. Salice, D. Sciuto: Con-

current error detection at architectural level, Proc.

11thInt.Symp.Syst.Synth.(1998)pp.72–75

30.10 C. Bolchini, L. Pomante, F. Salice, D. Sciuto: Re-

liability properties assessment at system level:

a co-design framework, J. Electron. Test. 18(3),

351–356 (2002)

30.11 M.D. Jeng: Petri nets for modeling automated

manufacturing systems with error recovery, IEEE

Trans. Robot. Autom. 13(5), 752–760 (1997)

30.12 G.A. Kanawati, V.S.S. Nair, N. Krishnamurthy,

J.A. Abraham: Evaluation of integrated system-

level checks for on-line error detection, Proc. IEEE

Int. Comput. Perform. Dependability Symp. (1996)

pp. 292–301

30.13 B.D. Klein: How do actuaries use data contain-

ing errors?: models of error detection and error

correction, Inf. Resour. Manag. J. 10(4), 27–36 (1997)

30.14 M. Ronsse, K. Bosschere: Non-intrusive detection

of synchronization errors using execution replay,

Autom. Softw. Eng. 9(1), 95–121 (2002)

30.15 O. Svenson, I. Salo: Latency and mode of error de-

tection in a process industry, Reliab. Eng. Syst. Saf.

73(1), 83–90 (2001)

30.16 X.W. Chen, S.Y. Nof: Prognostics and diagnostics of

conﬂicts and errors over e-Work networks, Proc.

19th Int. Conf. Production Research (2007)

30.17 J. Gertler: Fault Detection and Diagnosis in Engi-

neering Systems (Marcel Dekker, New York 1998)

30.18 M. Klein, C. Dellarocas: A knowledge-based

approach to handling exceptions in workﬂow sys-

tems, Comput. Support. Coop. Work 9, 399–412

(2000)

30.19 A. Raich, A. Cinar: Statistical process monitoring

and disturbance diagnosis in multivariable con-

tinuous processes, AIChE Journal

42(4), 995–1009

(1996)

30.20 C.-Y. Chang, J.-W. Chang, M.D. Jeng: An unsuper-

vised self-organizing neural network for automatic

semiconductor wafer defect inspection, IEEE Int.

Conf. Robot. Autom. ICRA (2005) pp. 3000–3005

30.21 M. Moganti, F. Ercal: Automatic PCB inspection sys-

tems, IEEE Potentials 14(3), 6–10 (1995)

30.22 H. Rau, C.-H. Wu: Automatic optical inspection

for detecting defects on printed circuit board in-

ner layers, Int. J. Adv. Manuf. Technol. 25(9–10),

940–946 (2005)

30.23 J.A. Calderon-Martinez, P. Campoy-Cervera: An

application of convolutional neural networks for

automatic inspection, IEEE Conf. Cybern. Intell.

Syst. (2006) pp. 1–6

30.24 F. Duarte, H. Arauio, A. Dourado: Automatic system

for dirt in pulp inspection using hierarchical image

segmentation, Comput. Ind. Eng. 37(1–2), 343–346

(1999)

30.25 J.C. Wilson, P.A. Berardo: Automatic inspection of

hazardous materials by mobile robot, Proc. IEEE

Int. Conf. Syst. Man. Cybern. 4, 3280–3285 (1995)

30.26 J.Y. Choi, H. Lim, B.-J. Yi: Semi-automatic pipeline

inspection robot systems, SICE-ICASE Int. Jt. Conf.

(2006) pp. 2266–2269

30.27 L.V. Finogenoy, A.V. Beloborodov, V.I. Ladygin,

Y.V. Chugui, N.G. Zagoruiko, S.Y. Gulvaevskii,

Y.S. Shul’man, P.I. Lavrenyuk, Y.V. Pimenov: An

optoelectronic system for automatic inspection of

Part C 30

Automating Errors and Conﬂicts Prognostics and Prevention References 523

the external view of fuel pellets, Russ. J. Nondestr.

Test. 43(10), 692–699 (2007)

30.28 C.W. Ni: Automatic inspection of the printing con-

tents of soft drink cans by image processing

analysis, Proc. SPIE 3652,86–93(2004)

30.29 J. Cai, G. Zhang, Z. Zhou: The application of

area-reconstruction operator in automatic vi-

sual inspection of quality control, Proc. World

Congr. Intell. Control Autom. (WCICA), Vol. 2 (2006)

pp. 10111–10115

30.30 O. Erne, T. Walz, A. Ettemeyer: Automatic shearog-

raphy inspection systems for aircraft components

in production, Proc. SPIE 3824, 326–328 (1999)

30.31 C.K. Huang, L.G. Wang, H.C. Tang, Y.S. Tarng: Auto-

matic laser inspection of outer diameter, run-out

taper of micro-drills, J. Mater. Process. Technol.

171(2), 306–313 (2006)

30.32 L.Chen,X.Wang,M.Suzuki,N.Yoshimura:Opti-

mizing the lighting in automatic inspection system

using Monte Carlo method, Jpn. J. Appl. Phys., Part

1 38(10), 6123–6129 (1999)

30.33 W.C. Godoi, R.R. da Silva, V. Swinka-Filho: Pattern

recognition in the automatic inspection of ﬂaws in

polymeric insulators, Insight Nondestr. Test. Cond.

Monit. 47(10), 608–614 (2005)

30.34 U.S. Khan, J. Igbal, M.A. Khan: Automatic inspec-

tion system using machine vision, Proc. 34th Appl.

Imag. Pattern Recognit. Workshop (2005) pp. 210–

215

30.35 L.H.Chiang,R.D.Braatz,E.Russell:Fault Detec-

tion and Diagnosis in Industrial Systems (Springer,

London New York 2001)

30.36 S. Deb, K.R. Pattipati, V. Raghavan, M. Shakeri,

R. Shrestha: Multi-signal ﬂow graphs: a novel ap-

proach for system testability analysis and fault

diagnosis, IEEE Aerosp. Electron. Syst. Mag. 10(5),

14–25 (1995)

30.37 K.R. Pattipati, M.G. Alexandridis: Application of

heuristic search and information theory to sequen-

tial fault diagnosis, IEEE Trans. Syst. Man. Cybern.

20(4), 872–887 (1990)

30.38 K.R. Pattipati, M. Dontamsetty: On a generalized

test sequencing problem, IEEE Trans. Syst. Man.

Cybern. 22(2), 392–396 (1992)

30.39 V. Raghavan, M. Shakeri, K. Pattipati: Optimal and

near-optimal test sequencing algorithms with re-

alistic test models, IEEE Trans. Syst. Man. Cybern. A

29(1), 11–26 (1999)

30.40 V. Raghavan, M. Shakeri, K. Pattipati: Test se-

quencing algorithms with unreliable tests, IEEE

Trans. Syst. Man. Cybern. A 29(4), 347–357 (1999)

30.41 M. Shakeri, K.R. Pattipati, V. Raghavan, A. Patter-

son-Hine, T. Kell: Sequential Test Strategies for

Multiple Fault Isolation (IEEE, Atlanta 1995)

30.42 M. Shakeri, V. Raghavan, K.R. Pattipati, A. Patter-

son-Hine: Sequential testing algorithms for multi-

ple fault diagnosis, IEEE Trans. Syst. Man. Cybern.

A 30(1), 1–14 (2000)

30.43 F. Tu, K. Pattipati, S. Deb, V.N. Malepati: Multiple

Fault Diagnosis in Graph-Based Systems (Inter-

national Society for Optical Engineering, Orlando

2002)

30.44 F. Tu, K.R. Pattipati: Rollout strategies for sequen-

tial fault diagnosis, IEEE Trans. Syst. Man. Cybern.

A 33(1), 86–99 (2003)

30.45 F.Tu,K.R.Pattipati,S.Deb,V.N.Malepati:Com-

putationally efﬁcient algorithms for multiple fault

diagnosis in large graph-based systems, IEEE Trans.

Syst. Man. Cybern. A 33(1), 73–85 (2003)

30.46 C. Feng, L.N. Bhuyan, F. Lombardi: Adaptive

system-level diagnosis for hypercube multiproces-

sors, IEEE Trans. Comput. 45(10), 1157–1170 (1996)

30.47 E.M. Clarke, O. Grumberg, D.A. Peled: Model Check-

ing (MIT Press, Cambridge 2000)

30.48 C. Karamanolis, D. Giannakopolou, J. Magee,

S. Wheather: Model checking of workﬂow schemas,

4th Int. Enterp. Distrib. Object Comp. Conf. (2000)

pp. 170–181

30.49 W.Chan,R.J.Anderson,P.Beame,D.Notkin,

D.H. Jones, W.E. Warner: Optimizing symbolic

model checking for state charts, IEEE Trans. Softw.

Eng. 27(2), 170–190 (2001)

30.50 D. Garlan, S. Khersonsky, J.S. Kim: Model check-

ing publish-subscribe systems, Proc. 10th Int. SPIN

Workshop Model Checking Softw. (2003)

30.51 J. Hatcliff, W. Deng, M. Dwyer, G. Jung,

V.P. Ranganath: Cadena: An integrated develop-

ment, analysis, and veriﬁcation environment for

component-based systems, Proc. 2003 Int. Conf.

Softw. Eng. (ICSE 2003) (Portland 2003)

30.52 T. Ball, S. Rajamani: Bebop: a symbolic mod-

elchecker for Boolean programs, Proc. 7th Int. SPIN

Workshop, Lect. Notes Comput. Sci. 1885, 113–130

(2000)

30.53 G. Brat, K. Havelund, S. Park, W. Visser: Java

PathFinder – a second generation of a Java model-

checker, Proc. Workshop Adv. Verif. (2000)

30.54 J.C. Corbett, M.B. Dwyer, J. Hatcliff, S. Laubach,

C.S. Pasareanu, Robby, H. Zheng: Bandera: Ex-

tracting ﬁnite-state models from Java source code,

Proc. 22nd Int. Conf. Softw. Eng. (2000)

30.55 P. Godefroid: Model-checking for programming

languages using VeriSoft, Proc. 24th ACM Symp.

Princ. Program. Lang. (POPL’97) (1997) pp. 174–186

30.56 Robby, M.B. Dwyer, J. Hatcliff: Bogor: An extensible

and highly-modular model checking framework,

Proc. 9th European Softw. Eng. Conf. held jointly

with the 11th ACM SIGSOFT Symp. Found. Softw. Eng.

(2003)

30.57 S. Mitra, E.J. McCluskey: Diversity techniques for

concurrent error detection, Proc. IEEE 2nd Int.

Symp. Qual. Electron. Des. IEEE Comput. Soc., 249–

250 (2001)

30.58 S.-L. Chung, C.-C. Wu, M. Jeng: Failure Diagnosis:

A Case Study on Modeling and Analysis by Petri Nets

(IEEE, Washington 2003)

Part C 30

524 Part C Automation Design: Theory, Elements, and Methods

30.59 P.S. Georgilakis, J.A. Katsigiannis, K.P. Valavanis,

A.T. Souﬂaris: A systematic stochastic Petri net

based methodology for transformer fault diagno-

sis and repair actions, J. Intell. Robot. Syst. Theory

Appl. 45(2), 181–201 (2006)

30.60 T. Ushio, I. Onishi, K. Okuda: Fault Detection Based

on Petri Net Models with Faulty Behaviors (IEEE, San

Diego 1998)

30.61 M. Rezai, M.R. Ito, P.D. Lawrence: Modeling and

Simulation of Hybrid Control Systems by Global

Petri Nets (IEEE, Seattle 1995)

30.62 M. Rezai, P.D. Lawrence, M.R. Ito: Analysis of Faults

in Hybrid Systems by Global Petri Nets (IEEE, Van-

couver 1995)

30.63 M. Rezai, P.D. Lawrence, M.B. Ito: Hybrid Modeling

and Simulation of Manufacturing Systems (IEEE,

Los Angeles 1997)

30.64 M. Sampath, R. Sengupta, S. Lafortune, K. Sin-

namohideen, D. Teneketzis: Diagnosability of

discrete-event systems, IEEE Trans. Autom. Control

40(9), 1555–1575 (1995)

30.65 S.H. Zad, R.H. Kwong, W.M. Wonham: Fault diag-

nosis in discrete-event systems: framework and

model reduction, IEEE Trans. Autom. Control 48(7),

1199–1212 (2003)

30.66 M. Zhou, F. DiCesare: Petri Net Synthesis for Discrete

Event Control of Manufacturing Systems (Kluwer,

Boston 1993)

30.67 Q. Wenbin, R. Kumar: Decentralized failure diag-

nosis of discrete event systems, IEEE Trans. Syst.

Man. Cybern. A 36(2), 384–395 (2006)

30.68 A. Brall: Human reliability issues in medical care -

a customer viewpoint, Proc. Annu. Reliab. Maint.

Symp. (2006) pp. 46–50

30.69 H. Furukawa: Challenge for preventing medication

errors-learn from errors-: what is the most ef-

fective label display to prevent medication error

for injectable drug? Proc. 12th Int. Conf. Hum.-

Comput. Interact.: HCI Intell. Multimodal Interact.

Environ., Lect. Notes Comput. Sci. 4553, 437–442

(2007)

30.70 G. Huang, G. Medlam, J. Lee, S. Billingsley, J.-

P. Bissonnette, J. Ringash, G. Kane, D.C. Hodgson:

Error in the delivery of radiation therapy: results

of a quality assurance review, Int. J. Radiat. Oncol.

Biol. Phys. 61(5), 1590–1595 (2005)

30.71 A.-S. Nyssen, A. Blavier: A study in anesthesia,

Ergonomics 49(5/6), 517–525 (2006)

30.72 K.T. Unruh, W. Pratt: Patients as actors: the

patient’s role in detecting, preventing, and recov-

ering from medical errors, Int. J. Med. Inform. 76(1),

236–244 (2007)

30.73 C.C. Chao, W.Y. Jen, M.C. Hung, Y.C. Li, Y.P. Chi:

An innovative mobile approach for patient safety

services: the case of a Taiwan health care provider,

Technovation 27(6–7), 342–361 (2007)

30.74 S. Malhotra, D. Jordan, E. Shortliffe, V.L. Patel:

Workﬂow modeling in critical care: piecing to-

gether your own puzzle, J. Biomed. Inform. 40(2),

81–92 (2007)

30.75 T.J. Morris, J. Pajak, F. Havlik, J. Kenyon,

D. Calcagni: Battleﬁeld medical information

system-tactical (BMIST): The application of mobile

computing technologies to support health surveil-

lance in the Department of Defense, Telemed. J.

e-Health 12(4), 409–416 (2006)

30.76 M. Rajendran, B.S. Dhillon: Human error in health

care systems: bibliography, Int. J. Reliab. Qual. Saf.

Eng. 10(1), 99–117 (2003)

30.77 S.Y. Nof: Design of effective e-Work: review of

models, tools, and emerging challenges, Product.

Plan. Control 14(8), 681–703 (2003)

30.78 X. Chen: Error detection and prediction agents

and their algorithms. M.S. Thesis (School of Indus-

trial Engineering, Purdue University, West Lafayette

2005)

30.79 X.W. Chen, S.Y. Nof: Error detection and prediction

algorithms: application in robotics, J. Intell. Robot.

Syst. 48(2), 225–252 (2007)

30.80 X.W. Chen, S.Y. Nof: Agent-based error prevention

algorithms, submitted to the IEEE Trans. Autom.

Sci. Eng. (2008)

30.81 K. Duffy: Safety for proﬁt: Building an error-

prevention culture, Ind. Eng. Mag. 9,41–45

(2008)

30.82 K.S.Barber,T.H.Liu,S.Ramaswamy:Conﬂictde-

tection during plan integration for multi-agent

systems, IEEE Trans. Syst. Man. Cybern. B 31(4),

616–628 (2001)

30.83 G.M.P. O’Hare, N. Jennings: Foundations of Dis-

tributed Artiﬁcial Intelligence (Wiley, New York

1996)

30.84 M. Zhou, F. DiCesare, A.A. Desrochers: A hybrid

methodology for synthesis of Petri net models for

manufacturing systems, IEEE Trans. Robot. Autom.

8(3), 350–361 (1992)

30.85 J.-Y. Shiau: A formalism for conﬂict detection and

resolution in a multi-agent system. Ph.D. Thesis

(Arizona State University, Arizona 2002)

30.86 J.A. Ceroni, A.A. Velásquez: Conﬂict detection and

resolution in distributed design, Prod. Plan. Con-

trol 14(8), 734–742 (2003)

30.87 T. Jiang, G.E. Nevill Jr: Conﬂict cause identiﬁca-

tion in web-based concurrent engineering design

system, Concurr. Eng. Res. Appl. 10(1), 15–26

(2002)

30.88 M.A. Lara, S.Y. Nof: Computer-supported conﬂict

resolution for collaborative facility designers, Int.

J. Prod. Res. 41(2), 207–233 (2003)

30.89 P. Anussornnitisarn, S.Y. Nof: The design of active

middleware for e-Work interactions, PRISM Res.

Memorandum (School of Industrial Engineering,

Purdue University, West Lafayette 2001)

30.90 P. Anussornnitisarn, S.Y. Nof: e-Work: the chal-

lenge of the next generation ERP systems, Prod.

Plan. Control 14(8), 753–765 (2003)

Part C 30