ML20134E679

Responds to Which Requested That Staff Respond in Writing to Questions 1,2,3 & 11,concerning Human Program Plan

ML20134E679
Person / Time
Issue date:	02/04/1997
From:	Thompson H NRC OFFICE OF THE EXECUTIVE DIRECTOR FOR OPERATIONS (EDO)
To:	Seale R Advisory Committee on Reactor Safeguards
References
ACRS-GENERAL, NUDOCS 9702060413
Download: ML20134E679 (9)

Text

~

February 4,1997 Dr. Robert L. Seale, Chairman Advisory Committee on Reactor Safeguards U.S. Nuclear Regulatory commission Washington, D.C. 20555 l

Dear Dr. Seale:

~

This responds to the letter from Dr. Kress dated December 30,1996, in which he asked the staff to respond in writing to four questions (Questions 1,2,3 and 11) l concerning its Human Performance Program Plan. The ACRS' questions and the staff's responses are enclosed. ,

' Sincerely, Origiddanedby H.as h L Thompson, #.

Hugh L. Thompson, Jr. i Acting Executive Director i for Operations

Enclosure:

As stated cc: Chairman Jacksor.

Commissioner Ro;ms -

1 Commissioner Dicus Commissioner Din Commissioner McGaffigan SECY Distribution:

HThompson EJordan FMiraglia DMorrison, RES Dross, AEOD CPaperiello, NMSS JBlaha AThadani BSheron RZimmerman TMartin JMitchell KCyr, OGC KBoher (G960977) WHodges FCoffman JPersensky ACRS File PDR HHFB RF

\ \

Document Naras: G:\SONARY\ACRSLTR & G:\ THOMAS \ACRSRESP L 03

'See orevious concurrence p [l 4 rp g ( Q $

Te receive a ospy of this ' 1 indiouse in the ben 'C' = Copy wr w h'encionwe 'E' = Copy wish hiencianus 'N' = No copy l pFFICE HHFB/DRCH l DRCH l , 'T l NRR l RES l EDQfL;.  !

y,'ff".E CTi.cmas:sc BBoger BSheron FMiraglia DMorrison HTh6rdpson 1/09,97* 1/15/97* 1/15/97" 1/17/97* 1/15/97* A ///97

{ ATE

. OFFICIAL RECORD COPY l

l c60cso NRC FILE CENTER COPY k l

i 9702060413 970204 PDR ACRS PDR

%, 7 i GENERAL l

, nad M-4 A 4 +4d 9&n- - 4Aw=,,p>3 # A A *\hMW%. 'b #' 'b v . .

f Dr. Thomas S. Kress, Chairman l

Advisory Committee on Reactor Safeguards -

U.S. Nuclear Regulatory commission Washington, D.C. 20555

Dear Dr. Kress:

This responds to your letter of December 30,1996,in which you ed the staff to respond in writing to four questions (Questions 1,2,3 and 11) conc ning its Human Performance Program Plan. The ACRS' questions and the staff' esponses are enclosed.

Sincerely, Hugh . Thompson, Jr.

Acti Executive Director i Operations

Enclosure:

As stated l

cc: Chairman Jackson Commissioner Rogers Commissioner Dicus Commissioner Diaz Commissioner McGaffigan SECY Distribution:

HThompson FMiraglia DMorrison, RES i T>.Ron lideedes, AEOD l CPaperiello, NMSS l JBlaha l

AThadani BSheron  !

RZimmerman TMartin JMitchel KCyr, OGC KBoher (G96 77)

ACRS File  !

PDR HHFB RF l Doc nt Name: G:\SONARY\ACRSLTR & G:\ THOMAS \ACRSRESP j T. . .t a. - i.si i. m. w c - c, .m m.a r h c4.m w a x - N. l

! OFFICE HHfB/DRC_H 6 DRCH l RES l 4Q)Wy l lRES l EDO l NAME ptptimet:sc BBoger "tr BSheronJg/,rrM@a DMorrison HThompson 2

DATE 1/tj97 1//5/97 1(5fG~i g)1/ 197 l 1 A6/97 / /97 OFFICIAL RECORD COP

Dr. Thomas S. Kress, Chairman 1 Advisory Committee on Reactor Safeguards l U.S. Nuclear Regulatory commission  ;

Washmgton, D.C. 20555 l

Dear Dr. Kress:

This responds to your letter of December 30,1996,in which you asked the staff to l respond in writing to four questions (Question 1,2,3 and 11) concerning its Human l Perfor- we Program Plan. The ACRS' questions and the staff's responses are enclosed.

Sincerely, i

l j Hugh L. Thompson, Jr.

j Acting Executive Director  !

j for Operations

Enclosure:

As stated i

3 cc: Chairman Jackson l

Commissioner Rogers d

Commissioner Dicus  !

Commissioner Diaz l

]I Commissioner McGaffigan i i SECY i

Distribution:

l HThompson i

FMiraglia

! DMorrison, RES j EJordan, AEOD j CPaperiello, NMSS i JBlaha 3

AThadani 4

BSheron <

', RZimmerman 1 TMartin l

! JMitchel KCyr, OGC KBoher (G960977)

ACRS File  !

PDR HHFB RF _

Document Name: G:\SONARY\ACRSLTR & G:\ THOMAS \ACRSRESP T. w...w.rei. - wwm.we.w c-c., rom - w r-c.,y.m e .-w

)

v-w.

pFFICE HHFB/DRCH DRCH RES ADT ~RES fl' EDO l PAME CThomas:sc BBoger . BSheron TMiraglia DMorrisfon HThompson pATE 1/ /97 1/ /97 1/ /97 if /97 1/1797 / /97 OFFIClhL RECORD COPY

. )

ENCLOSURE RESPONSES TO ACRS' OUESTIONS ON THE HUMAN PERFORMANCE PROGRAM PLAN Question 1. What are the staff plans for developing a Human Performance  ;

Program Plan (HPPP) activities road map, which would be useful for allocating resources, scheduling, and understanding the relationship between the activities?

Resporae: The HPPP was intended to function as a focal point for the coordination of human factors activities among the NRC offices.

Planning and implementation of these activities, including allocation of resources, is the responsibility of the individual offices.

The HPPP describes the mission, goals, objectives and activities of the Agency's human performance program. The HPPP's goals and objectives are aimed at assuring that (1) an acceptable level of human performance is maintained at nuclear facilities, (2) sufficient information is available to support the regulation of human ,

performance and (3) adequate resources are available to carry out its Sman performance program. The HPPP also describes the activities tnat are necessary to accomplish the goals and objectives, the office that is responsible for accomplishing each activity, the schedule for completing each activity, the relative priority of each activity and the relationships between and among the various activities.

l The framework for the HPPP was developed at a facilitated workshop i using a panel of internal human performance experts, some from the  !

former Human Factors Coordination Committee. The panel started )

from a simplified organizational systems model and used its extensive

. education and experience to develop ths mission statement, goals

( and objectives of the HPPP. It then developed the activities necessary to support the HPPP's goals and objectives. A "roadmap" depicting the relationships between and among the various elements  !

l of the HPPP is shown in Figure 1.  ;

We find that the HPPP adequately serves its intended purpose-to function as a focal point for the coordination of tha Agency's human factors activities. An additional HPPP roadmap developed from a human performance modelis not needed and cannot be justified, especially in light of the Agency's higher priority human factors work and limited human factors resources. Nevertheless, it ie important to l

note that the staff does have a cr.,mmitment to periodically consider the need to update and refine the HPPP.

1 4

FIGURE 1

,r 1r 1r inspection Experience from

^ Experience Other Settings a e

I2.2.1-2.2.5 r- 1 y rs 1.2.10,2.1.2

........,..r m'"n >

sj *

^ 3' . '1. 5 u

3 NO issue l Identfication 2.2.8,2.2.9,2.4.5 J

.: p w 1r

+YES / /s Guidance I

\

N NO /-'N Are Resource s \

Available? YES Available?

+ . .

, 3.3.1,3.2.1 -3.2.3 Develop \

NO / Guidance?

% Policy DecIslon, b/ YES HPPP

• x Roadmap ufficien l YES .

/ Information NO IE@ Available?  !

Prepare -

3r Guidance I1.3.1,2.1.1, 2.1.6 2

Develop Technical Basis 1.1.6-1.1.8,1.2.4-1.2.9, I #

1.3.2,1.3.4,2.4 r +- ~s implement I Guidance 1.1.2,1.2.2,1.2.3, I

1.2.10,2.1.2 s

y: "Q

l l Question 2. The activities delineated in the HPPP appear to be focused on I

reducing the assumed risk-worth of human actions used in probabilistic risk assessments (PRAs). What is the risk-worth of human actions? Why does the staff believe the risk-worth is too high and should be reduced?"

Response: The HPPP was built upon feedback from experience and the staff's

, expertise accrued from studying human performance issues over i time, both within the nuclear industry and from other applications, I

a.g., military, aerospace, transportation. Feedback from operating experience demonstrates that more than half of the incidents reported for both nuclear power plants and nuclear materials facilities have human performance as a contributing factor. Associated with this i data is the staff's experience with the risk focus of the agency.

Though there was no explicit attempt to apply quantitative risk worth to the activities, one of the premises on which the HPPP was developed (as delineated in the August 1995 version of the HPPP) was that " human error is a significant contributer to risk associated i

with operational events at nuclear facilities." l l

Several studies allude to the " risk worth of human actions used in i PRAs." The early Brookhaven National Laboratory (BNL) sensitivity studies (NUREG/CR-1879) and the S-curves BNL generated for a PWR, for example, showed that the core damage frequency (CDF) is significantly increased by an increase in human-error rates. The Individual Plant Examinetion (IPE) draft report, NUREG-1560, lists BWR end PWR scenarios in which human actions are important and categorizes them by percentage of the plants at which they are important, as follows: " Human actions arc importam contributors to plant CDFs in the IPEs, with correct operator actions often significantly seducing the CDFs. However, there are considerable uncerte!nties associated with determining human error probabilities for operator actions. As a result, improved modeling of human actions would significantly improve the understanding of the risk associated with operating nuclear power plants." Therefore, human actions may increase or reduce the level of risk.

Determining " risk-worth of human actions in PRAs" requires that i

human actions be converted into some risk measure. Past methods l by which human actions are converted into some risk measure involved large uncertainties in determining the risk-worth of human l actions. The methods appear incomplete, particularly in modeling j errors of commission and cognition. It is premature to establish quantitative values for " risk-worth" of human actions before j developing a more complete method to convert human actions into a i risk measure. Therefore, feedback from operating experience and the j collective experience and expertise of those formulating the HPPP 1

was the primary basis for the HPPP. The premises that served as the i basis of the original HPPP are listed in the August 1995 version.

i

l .. .

Ouestion 3. How does the staff set the priorities for the HPPP activities and what does the priority ranking mean?

Response: The actual or effective priorities of each of the HPPP's activities are determined by the office director that is responsible for completing the activity in the context of that office's total workload using its -

own work prioritization scheme. The priorities listed in the HPPP are considered in the office director's priority determination.

The priorities listed in the HPPP - high, medium and low - represent the former Human Factors Coordination Committee's qualitative assessment of each activity's relative importance, taking into consideration perceived safety significance, immediacy, management i direction and regulatory need. The HPPP priorities were determined by consensus from the Committee's members and, thereby, indirectly reflect the perspective of each of the offices represented by the Committee meffJership. The activities' priorities are expected to change as their underlying considerations change. The priorities of each of the activities will be reconsidered periodically,'as will the ,

need to update the HPPP.  !

i l

i i

i i r l

l l

l  :

l Question 11. How are etandards adopted by the staff formulated? How does the staff assure that the standards are necessary and sufficient to meet ,

! regulatory needs?

[

t l i Response: Human factors consensus standards are primarily based on the  !

l accrual of evidence from experience. They are used as appropriate in lieu of expending substantial resources to develop an experimental i base. The limitations on experimental evidence do not preclude a technical basis for standards. It is important to check that the scope l of the consensus standard should not exceed the scope of the l

experience base. Human factors consensus standards are produced I through an open and mature process by national standards l l

organizations composed of working groups interested in specific j topics. General practice is to reassess these standards approximately I every 5 years. l l

The NRC's use of consensus standards is consistent with the policy established for Executive Branch agencies by OMB Circular No. A-119, Revised, October 20,1993. That circular states,in part, j_ lt is the policy of the Federal Government in its procurement and regulatory activities to: Rely on voluntary standards, both domestic and international, whenever feasible and consistent with law and '

regulation pursuant to law ....

NRC Direction Setting issue 13 encourages the use of consensus standards in stating that The codes and standards that are established and referenced by NRC incorporate many years of accepted, good engineering practice and are relied on to provide a rational basis for many NRC regulations.

Standards are selected for adoption by the NRC based on their relevance to the regulatory process. The human factors standards which are used currently by the NRC staff relate to personnel and 1 operational issues, e.g., training and qualifications, medical i qualification, simulator fidelity, procedures, system interface. These areas are cited in the regulations or the Standard Review Plan. ,

j NUREG-0800 (SRP). The standards are more detailed than the l l regulations or SRP. This level of detailis needed by the regulatory j staff to use as part of their evaluation of licensee submittals or for l

inspections. The standards are selected with the intent of endorsing 1 i

those that are sufficient to meet the Agency's primary regulatory l need - maintaining the safety of nuclear facilities.

i '

Consensus standards are endorsed by regulatory guides that are not j

I

4 mandatory for licensees. Regulatory guides do not prescribe requirements: they only describe methods that the staff has determined to be acceptable ways of meeting the mandatory regulations.

Before the NRC endorses a consensus standard with a regulatory guide, there are additional checks and balances that accrue. The NRC can, and does, take exception to certain elements of the standard on the basis of Agency experience. There is also a mandatory public l comment period, during which anyone can make comments on the document that must be considered and acted on by the staff. Finally, 1 various levels of management, the CRGR, and the ACRS review the document.

i Much of the NRC human factors guidance extant is not in the form of 1 i

regulatory guides, but is promulgated as staff reports, e.g., i NUREG-0700, NUREG-0899. These documents do not endorse standards in the same way a regulatory guide does, but incorporate l guidance that has been adapted from standards developed for other l applications, e.g., military, aerospace, transportation, industrial safety. Before these standards are incorporated in NRC guidance, efforts are made to determine the basis for the elaments of the standards. The staff has followed a review process that determines (1) if the standard is based on laboratory or field research or

experience, (2) the history of use of the standard in its intended I l application, and (3) the reasonableness of its fit in a nuclear

! application. These documents are frequerstly peer reviewed by l subject matter experts and may be subject to public comment and CRGR and ACRS review.

r i

4 4

t .

du uouq y 't* UNITED STATES NUCLEAR REGULATORY COMMISSION f WASHINGTON, D.C. 2056tHX101

%.....$ February 4,1997 Dr. Robert L. Seale, Chairman Advisory Committee on Reactor Safeguards U.S. Nuclear Regulatory commission Washington, D.C. 20555

SUBJECT:

ACRS QUESTIONS ON HUMAN PERFORMANCE PROGRAM PLAN This responds to the k$tter from Dr. Kress dated December 30,1996, in which he l

asked the staff to respond in writing to four questions (Questions 1,2,3 and 11) 1 conceming its Human Performance Program Plan. The ACRS' questions and the staff's responses are enclosed.

Sincerely, i i

l

. < A?r l HL gh L. Thompsony /J r l Ac g Executive re t6r j for Operations l

Enclosure:

As stated i i

cc: Chairman Jackson Commissioner Rogers Commissioner Dicus i Commissioner Diaz Commissioner McGaffigan SECY l l t l l

l

.e_

. I g

. .. ~

P. , ,G7[l f) b E V , ,

U d h[ME!li!fid$3 EDO Principal Correspondence Control FROM: DUE: 01/22/97 EDO CONTROL: G960977 DOC DT: 12/30/96 FINAL REPLY:

T. S. Kress ACRS

.TO:

James M. Taylor FOR SIGNATURE OF : ** GRN ** CRC NO:

Exncutive Director DESC: ROUTING:

ACRS QUESTIONS ON HUMAN PERFORMANCE PROGRAM PLAN Taylor Milhoan

'- Thompson

~

Blaha Mitchell Paperiello,NMSS DATE: 12/31/96 c Morrison,RES

' Jordan,AEOD ASSIGNED TO: CONTACT: A Cyr, OGC ACRS File w

NRR Miraglia SPECIAL INSTRUCTIONS OR REMARKS:

Prnpare response to ACRS for EDO signature.

Put Commissioners and SECY on cc (shown on F IN original) for reply.

]4 h f /3 lfh USE SUBJECT LINE IN RESPONSE.

~

NRR RECEIVED: DECEMBER 31, 1996 .

NRR ACTION: DRCH:B0GER AC,il0N NRR ROUTING: MIRAGl_IA THADANI j)iJri TO i s IX',EGOR'S Ci!:lCE I

ZIMMERMAN SHER 0M wq gf / +6/ C[ Cf' 7~~

TRAVERS ,

- ~ ~ ~

PARTIN

~

B0HRER l

v-

• p,# U:tuq4 e UNITED STATES

. 8* $ NUCLEAR HEGULATORY COMMISSION

$ ,$ ADVISORY COMMITTEE ON REACTOR SAFEGUARDS

$ # WASHINGTON, D. C. 20555

'*sss*

December 30, 1996 Mr. James M. Taylor Executive Director for Operations U.S. Nuclear Regulatory Commission Washington, D.C. 20555-0001

SUBJECT:

ACRS QUESTIONS ON HUMAN PERFORMANCE PROGRAM PLAN During the 437th meeting of the Advisory Committee on Reactor Safeguards, December 5-7, 1996, we reviewed the NRC activities identified in the Human Performance Program Plan. Our Subcommittee on Human Factors met on September 20 and December 3, 1996, to review these activities. Af ter the September 20, 1996 Subcommittee meeting, a list of questions included in the attachment was developed. These questions were provided to the staff on September 27, 1996. During subsequent meetings, the staff responded to these questions. We believe that the staff's response to questions 1, 2, 3, and 11, did not fully address our concerns. We request that the staff provide written response to these questions.

Sincerely,

.J 5 /47 '

T. S. Kress Chairman

Attachment:

List of ACRS questions on Human Performance Program Plan cc: J. Mitchell, OEDO F. Miragila, NRR B. Boger, NRR C. Thomas, NRR D. Morrison, RES W. Hodges, RES 1

EDO -- G960977 I i

f -

.g,

J ATTACMENT 1 l

LIST 0F ACRS QUESTIONS ON NUMAN PERFORMANCE PR0GRAN PLAN i

The ACRS requested ' that the staff provide information at a future'ACRS Subcommittee meeting concerning the following questians.

1. What are the staff plans for developing a Human Performance Progr&m Plan (HPPP) activities road map, which would be useful for allocating resources, scheduling, and understanding the relationship between the .

. activities? l

2. The activities delineated in the HPPP appear to be focused on reducing the assumed risk-worth of human actions used in probabilistic risk assessments (PRA). What is the risk-worth of human actions? Why does the staff believe the risk-worth is too high and should be reduced?

3. How does the staff set the priorities for the HPPP activities and what does the priority ranking mean?

4. How does the staff decide that an independent program element is required?

Why has the staff decided that data gathering should be separated from developing guioance and ' that the two activities should have different priorities?

[ NOTE: The attached figures are examples of models that may be used to develop a master diagram that could serve as the road map to answer many of the questions raised here. These figures are just the starting point; they must be adapted to the NRC's needs using judgment and operational experience.)

5. What does the staff mean by " effective" and

• adequate" as used in the objectives and goals in the HPPP?- How does the staff know what must be done and when the goal or objective is achieved?

6. Should the staff be pushing licensees toward the state-of-the-art in human i factors and human reliability rather than a proven adequate state? l l

7. Numerous human errors have resulted in the misadministration of medical  !

treatments by licensees of the Office of Nuclear Materials' Safety and l Safeguards (NMSS). Why isn't NMSS as involved with human performance  !

efforts as the other offices?

i 8. How does the staff plan to respond to the ACRS advice concerning

' developing metrics for organizations and managements that correlate with  !

j risk or performance?

I l g. What are the technical bases for defining the staffing levels inside and  !

outside of the main control room, and for communication procedures? l t

j 10. What are the deficiencies or " holes" in NUREG-0700? .

_ .., ~ .-

W b w

. _._ _ _ ~ _ . _ - . . _ _ - - - .

[

. ~

l ATTACHMENT 1 l

i 11. How are standards adopted by the staff formulated? How does the staff l assure that the standards are necessary and sufficient to meet regulatory needs?

l 12. The staff scheduled item 1.2.11 of the HPPP, " Develop Guidance for Computerized Job Performance Aids," to be completed "as technology is developed." What standards does the staff have for such aids that would foster the development of such technology? If the standards do not exist, what are the staff plans for developing such standards?

13. What is the staff approach to developing a performance-based fitness-for-duty criteria?  !

l

14. What is the staff approach to evaluating the task network model espoused by the Department of Defense, and how will the staff decide if tne model  !

is applicable and useful for regulatory needs?

15. How does the staff decide on the allocation of resources between human I factor research and other research activities such as thermal hydraulic )

l models? l 1

16. How does the staff assure sirulator fidelity? How important is good fidelity to Emergency Operati+c Procedure training? What does the staff expect an operator to do if unexpected plant behavior occurs during a severe accident?

l I

l i

l I 1 i

\

1 W

4 1

m anc x m

l TYPES l AReflonelfuIUrDe j s teunen i cmionen j;$UP _

swrut .

m

,, UNWTENDED esseeg.

ACTION Memorylessee

. LAPSE -~. O mM

• W hans mee.*% '

UN$AFE ACT8 1 lhJo-basedtrdstakes / '

no ppacamenare end i J . AppkeenorIm6nde. l l h__ ggy m M

m.,M MOED . Manywriable twms.

ACTION ,

Routine violadons

--. Vict. Ail 0N ~ Exceptiond m Acts of sabotage Figur'e 7.7. A summary of the psychological varieties of un.

safe acts, classified initially according to *bether the act was intended or unintended and then distinguishing errors from violations. I j L'

. \

l F4anctms rana;Metoisus l I I so m runenen unes. l w w= l consm.an

w. won. .

uwE rsymouxWCAL rggg;7g, FALUBLE DEQ3 IONS "

Em "'

PRECURsoMS

• g l DERQEMEg of UNWE Acr8

, < 2 acts . <

NQoDtra >

1 l _

i '"" I toes ca.am Lang: upon adequacy eranow Loops -de8ences I t. cap i Figure 7.9. Feedback loops and indicators. The indicators are divided into two groups: failure oper (relatir.g to deficiencies in the managerial /organisational sectors) andfagure tokens (relating to indMdual conditions and unsafe acts).

J. Reason, Human Error, Cambridge University Press, 1990.

.. ,- _]

2: -

fa s

n M

w 3 a a y aW*

1

Error  ::

-  :: PRA .

Forc.mg .

:: Human Error  : :  : Model

c. Context  ::

=. .  ::

N NN"' .

Performance  !! Error  !  !

O s  !:

• Unsafe Sr ymg Factors M h .is m s Actions y ng i Mantensoce j jj j .

u.

s w n

.i  :  : e

.p g .

. soonerio j

, . DeAnition ,

l 1

I I

size

.. . . _ . ._..m

. _ _ _ _ ...~m. .... - . _ .. . ~ . _ . _ . .- ..

o. .

r e .

.. i 4

Factors affecting.performenee siteetion fasters Persempel task '

Subjective goals and intentions Task cherectoristice Mental load and resources Etymipment desigtt Physical environment Procedure design Affective factors Work time cheructoristics Fabrication Installation Inspection Operation Test and calibratioi keintenance, repe:r Lectatics Administration

)y 3 p Management

• Nechanisme of heutet matfswttoa Discristaetion

• 8tereotype fixation *

• Familiar shortcut 1y

• stereotype takeewer

• Familiar pottern not *nternal 3Musen , p I

recognized malfenetton 3

causes of heman malfsmeties suternal mode of selfumettom Input information processissg. Detection External events.

• Information not recalved Identification Specified task not performed distractions, etc*

• Mininterpretation Decision i

Excessive task demands,

• Assumption

• select goal

• Omission of act 4 4 *

• Inaccurate performance force, time, knowledge, etc. pece11 e select target

• Wrong timing Operator incapacitated.

• select task i sickness, etc.

• Forget isolated itose Action Commissions of erroneone act Intrinsic human verlebility

• Mistake alterwetives

• Operational coraission of extraneous act l

i

• Other slip of memory sequence Sneak-patha eccidental timing

(

• Etecution of sovsrel events or faults Inference e communication gl

• condition not considered l

• Side effect not considered..

Physical coordination

• Motor variability I

• Spatial miserientation i

Figure 11.2. Multifaceted taxonomy for description and analysis of events involving human malfunction. [ Reproduced from Rasmussen .(1982) with permission from John Wiley & Sons, Ltd.] g us l

J. Rasmussen, Information Processing and Human-Machine Interaction, North-Holland, 1986.

l i

l . _ - - _ _ .-

__ --- _ __ _ _ _ _ _ _ _ . =