HANDBOOK OF INDUSTRIAL ENGINEERING Edited by GAVRIEL-SALVENDY Purdue UniversityCopyright © 1982 by Johe Wiley & Sons, In. Al rights reserved. Published simultaneously in Canad Reproduction or translation of aay part of this work ‘beyond that permitted by Section 197 oF 108 of h 1976 United States Copynght Act without the permission ofthe copyright owner i unlawful. Requests for permission of further information should be addressed to the Permissions Department, John Wiley & Sons, Ine. Library of Congress Cataloging in Publication Data: Main entry under tite Handbook of industrial engineering. “A Wiley-Intericience publication.” Includes index 1. Industral engineering Handbooks, mits, et 1, Salvendy, Gavriel, 1938 T5623 36 65S 81-23059 ISBN 0.471.05881-6 ‘AACR Printed in the United States of Ameria Wo987654321 Virgil Profe Depa: Purdi West Thom Assist: Hy The U Ann Jeffre: Assist Depar Orego Corval Guy J Mana: Indust Intern Chicay Andre Profes Unive Minn: Most Profs Scho Purd West Fran Priva AnnCONTENTS INDUSTRIAL ENGINEERING FUNCTION 1.1 The Industrial Engineering Profession, by Byron W. Saunders 1.2 Organization and Administration of Industrial Engineering, by 0. J. Feorene 13 Improving the Effectiveness of Industrial Engineering Practice, by Gerald Nadler 14 Productivity: An Overview, by D. C. Burnham 15 Productivity Measurement and Improvement, by Marvin E. Mundet 16 Creative Problem Solving, by Knut Holt ORGANIZATION AND JOB DESIGN 2.1 Organization Design, by Louis E. Davis 22 Nonfinancial Motivation: Creating a Work Environment for Effective Human Performance, by James A. Richardson and Roy F. Lomicka 23. Financial Motivation, by Mitchell Fein 24 Job and Task Analysis, by Ernest J. McCormick 25 Job Design, by Louis E. Davis and Gerald J. Wacker METHODS ENGINEERING 3.1 Methods Design, by Charles E. Geisel 3.2 Motion Study, by Daniel O. Clark and Guy C. 33 Charting Techniques, by Takeji Kadota 34 Assembly Line Balancing, by Colin L. Moodie 3.5. Machine Interference: Assignment of Machines to Operators, by Kathryn E. Stecke PERFORMANCE MEASUREMENT AND CONTROL OF OPERATION 4.1 Work Standards: Establishment, Documentation, Usage, and Maintenance, by Joseph 4. Panico 42° Measuring and Controlling Machine Performance, by Alfred H. Smith and Irvin Otis 43° The Learning Curve, by Walton M. Hancock and Franklin H. Bayha Close, Jr. 1-28 117 2 128 1413 rer 1-40 a Lt ru tad 13 11s a3“i CONTENTS 44 Time Study, by Benjamin W. Niebel 45 Predetermined Motion Time Systems, by Chester L. Brisley and Karl Eady 46 Work Sampling, by W. J. Richardson and Eleanor S. Pape 4.7 Computerized Work Measurement, by Davendra Mishra 48 Development and Use of Standard Data, by Adam W. Cywar 4.9 Indirect Operations: Measurement and Control, by Guy J. Bacei 5 EVALUATION, APPRAISAL, AND MANAGEMENT OF HUMAN RESOURCES 5.1 Subjective Aspects of Performance, by Richard G, Pearson $2 Selection and Training of Personnel, by Walter C. Borman and Norman G. Peterson 53 Job Evaluation, by Ernest J. McCormick SA Performance Appraisal, by Irvin Otis and Robert W. Burns $8 Human Resources Accounting: Measurement and Utilization, ‘by Baruch Lev 5.6 Labor Relations: The Special Problems of the Industrial Engineer, by William Gomberg 6 ERGONOMICS/HUMAN FACTORS 6.1 Psychomotor Work Capabilities, by Gavri James L. Knight 62 Reduction of Human Error, by David Meister 63 Engineering Anthropometry and Occupational by Don B. Chaffin 64 Physiological Basis for the Design of Work and Rest, by Eliezer E. Kamon 65 Bodily Discomfort, by likka Kuorinka 6.6 Occupational Stress, by Gavriel Salvendy and Joseph Sharit 6.7 Work Performance and Handicapped Persons, by Thomas J. Armstrong and Dev S. Kochhar 68 Design of Man-Machine Systems, by Tarald O. Kvalseth 69 Design of Handtools, Machines, and Workplaces, by E. Nigel Corlett 6.10 Noise and Vibration, by Malcolm J. Crocker 11 Lighting, by Corwin A, Bennett 6.12 Climate, by Stephan Konz 6.13 Toxicology, by Kari Lindstrom 6.14 The Management of Occupational Safety, by James M. Miller vendy and jiomechanics, 7 MANUFACTURING ENGINEERING 7 Manufacturing Engineering, by Harold N. Bogart 7.2 Designing for Manufacturing, by Benjamin W. Niebel Lar 131 La 17 Lo con urents ‘CONTENTS: ail 7 13. Value Engineering, by David J. DeMarle and M. Larry Shillto 1-20 _ 14 Conventional Machine Tools, by Kenneth M. Gettelman 1 - 1S Automation, by Ralph E. Cross, Sr. M4 1.6 Industrial Robotics, by Ronald L. Tarvin, Merton D. Corwin, 12 and Wayne E. Mechlin Mt Dad 7.7 Numerical Control Machines, by Kenneth M. Gettelman 19 18 Group Technology, by Inyong Ham 1419 C3 79 Computerized Manufacturing Systems for Discrete Products, by Moshe M. Barash 7 7.40 Computers in Continuous Process Control, by _ Edward J. Kompass 8 TAL Sensor Technology, by Richard 4. Mathias 16 hat 8 QUALITY ASSURANCE 17 81 The Quality Assurance System, by Christian H. Gudnason 1-17 9 8.2 Measurement Assurance, by Karl F. Speiel 1-16 83 Quality Control, by Philip S. Brumbaugh 1-23 7 84. Improving Inspection Performance, by Colin G. Drury 14 : 85 Reliability and Maintainability, by Kailash C. Kapur 1 ” 9. ENGINEERING ECONOMY 9.1 Company and Cost Accounting, by Evan F. Bornkolts 1-20 92 Cost Estimating, by Phillip F. Ostwald 128 las 93 Discounted Cash Flow Techniques, by Raymond P. Lutz 1-25 9 9A Project Selection and Analysis, by Gerald J. Thuesen 127 i 95 Economic Risk Analysis, by James R. Buck and a Jose M. A. Tanchoco 1a 7 9.6 Production Economics, by Ferdinand F. Leimkuhler 17 is 10 FACILITIES DESIGN 1s 10.1 Location: Single and Multiple Facilities, by George M. Parks 1-14 10.2 Plant Layout, by James A. Tompkins 1a cs 103 Materials Handling Systems, by E. Ralph Sims, Jr 1-16 19 104 Storage and Warehousing, by John A. White and Hugh D. Kinney 1-32 ord 10.5 Energy Management, by Wayne C. Turner and “19 William J. Kennedy, Jr 1a 8 106 Office Layout, by John J. Mariott 1-20 110 11 PLANNING AND CONTROL . 1.1 Technological Forecasting, by David J. DeMarle and M. Larry Shillito 1S 11.2 Planning and Control for Manufacturing Systems and Projects, by Randall P. Sadowski and Deborah J. Medeiros 1-24 ie 113 Aggregate Production Scheduling, by Samuel Eilon 1-23 ie 114 Inventory Management and Control, by Urban Wemmerlév 1-28n CONTENTS Capacity: Its Measurement and Management, by Robert A. Boehmer Material Requirements Planning, by Gene J. D'Ovidio and Richard L. Behling Maintenance Management and Control, by Rintaro Muramatsu, Yoshihiko Tanaka, and Seiichi Nakajima. 1-19 118 Work Schedules, by Stanley D. Nollen 1a 119 Personnel Scheduling, by William J. Burgess and Robert E. Busby 16 M40" Purchasing: A Strategic Force, by Leroy H. Wullmeier and Michael R. Hottinger 113 ILI Distribution and Logistics, by John J. Jarvis and H. Donald Ratify 19 COMPUTERS AND INFORMATION PROCESSING SYSTEMS. 12.1 Computer Fundamentals, by Herbert D, Schwetman 116 122 Fundamentals of Information Processing Syste Classification and Characteristics, by Daniel Teichroew 16 123 Data Base Management, by Clyde Holsapple, Sheldon Shen, and Andrew B. Whinston 1 124 Analysis and Design Techniques for Information Processing Systems, by John V. Pilisis 19 125 Information Processing Systems and Computers in Industrial Engineering, by Hamed K. Eldin 0 126 Systems Design to Encourage Use, by Henry C. Lucas Jn 1.9 127 Office Automation, by Andrew D. Bailey, Jr, James Gerlach, R. Preston McAfee, and Andrew B. Whinston 1-20 QUANTITATIVE METHODOLOGIES FOR INDUSTRIAL ENGINEERS 13.1 Mathematics for Industrial Engineers, by Wiliam E. Biles tt 132 Concepts of Probability, by Robert G. Morris 19 133 Decision Principles and Anaiysis, by James R. Buck 17 134 Design of Experiments for Industrial Engineers, by Virgil L. Anderson and Robert 4. McLean 19 135 Hypothesis Testing and Statistical Inference, by Don T. Phillips 120 13.6 Regression and Correlation, by Douglas C. Crocker 12 13.7 Queuing Theory, by M. Raghavachari 1s 138 Markov Chains, by M. Raghavachari 0 13.9 Time Series Forecasting, by Arnold L. Sweet 1416 1340 Control Models, by Shimon ¥. Nof and Theodore J. Williams 1a 1B Computer Simulation, by James R. Wilson and A. Alan B. Pritsker coni 4 ¢ IndexCONTENTS 14. OPTIMIZATION IN INDUSTRIAL ENGINEERING Optimization: An Overview, by A. Ravindran and Gintaras V. Reklaitis Linear Programming, by A. Ravindran Nonlinear Optimization, by Katta G. Murty Discrete Optimization, by Jeffrey L. Arthur, James P. Caie, Jr., Joel D. Cohen, and Robert H. Harder Network Optimization, by Gary E. WhitehouseCHAPTER 12.7 Office Automation ANDREW D. BAILEY, JR. University of Minnesota JAMES GERLACH, R. PRESTON MCAFEE, ANDREW B. WHINSTON Pur Je University 12.7.1 INTRODUCTION Office automation is an evolving trend centered around the automation of office procedures and document manipulation by means of computer technology. The purpose of Section 127.2 is to ssquaint the reader with the generally accepted concepts and capabilites of office information Systems (DISe, the jses surrounding her evelopment, a ealfe prototype (Citibunk's Manage- ment Work Stations. and some of the implications tha office automation wil have for business And society ‘Section 12.7.3 presents two research OIS prototypes: Officetalk-Zero™ and System for Com- puterizstion of Oitice Proceses (SCOOP), OfficetalkZero ie an electronic ald to performing Foutine oftice tasks, similar In manner tothe traditional way of performing the equivalent clerical tasks; SCOOP stresies the utomation not ony of office devices, But also of office procedures. The need for nontechnical office personnel to implement automated office procedures leads toa di fussion of the office workeremachine interface in Section 12.7.4. Two potential languages for describing automated office procedures are presented: the Office Procedures by Example language find the Business Definition Language. Lach language is examined according 10 design philosophy, ‘objectives, prospective uses, and limitations Automated office procedures and devices are effective means by which the office worker can manipulate, store, and send electronic documents. Such capabilities should be adequate for sup- Porting formal communications. However, informal communications, serving important orpaniza- Fional needs, may not be adequately aesounted for by the document-oriented O1S. Thus Section 12.74 concludes with a discussion of formal and informal communications in an automated office Section 12.75 deals with the modeling of OISs in teuns of formal office activities and formal lines of cotnsumieation. A model for desribing and analyzing information flows within offices is presented. The model lenge Hell to automatic analysis and tansformations intended to automate, Sreamine, and reorganize the O1S model. The outlook for OIS simulation as an analyte tol and 2 petsonnel training tool follows. in addition to office ellicieacy and productivity, the office de Signer and pianoges rnost incorporate intesnal control mechanisms inthe automated office. A com Duterassisted infernal control verfiation system, TICOM I, is introduced as an effective means of Enalyzing OS models for violations of sound internal control. Finally, Section 12.7.6 teats the problem of preparing Yor the implementation of OSs. The chapter ends with some thoughts on the Inplieations of 01S for future office activities. This project was funded i part by a grant trom the Peat, Marwick, Mitchell Foundation through Its reveareh opportunities in auditing program. The views expressed herein are those of the authors tnd go not necsssrly reflect the views of the Peat, Marwick, Mitchell Foundation. The sesearch tras supported in part By Army Contract No. DATSCO1S4, We gratefully acknowledge the fellow Ship support made possible by a grant from IBMnaa COMPUTERS AND INFORMATION PROCESSING SYSTEMS 12.7.2 OFFICE INFORMATION SYSTEMS IN PERSPECTIVE With the introduction of the computer in the business environment, business functions such asin. ventory control, accounts receivable, accounts payable, and payroll accounting have boca sucess: fully automated. The benefis have included faster and mote accurate reporting, more ctfecive ‘Morage and management of lage amounts of data, and greater relabilty and conssiency of system, performance, However, desite the merits of automation, traditional office work has remained Felatvely unchanged. The goal of the automated OLS i to facilitate ofice workers performance of thr routine duties by providing them a tally officeoriented system that exploits the processing speed, precision, and generally infalible memory of eomputing machines, The OS is ifferentiated from an information processing system (IPS), 2 word procesie (WP), of 2 devision support sysicm (DSS) by its emphasis onthe integration of all of the eomponcnts nese sary to support office personncl. The IPS, WP, and DSS ar either nt ofiented toward otTcs Tune tions or addres only lintel set of office activities. The base functions of oftice work include text editing, document preparation and fing, copying simple numerical processing, communica ‘ion, and information analysis and verification, With respect to office procedures, a'WP and over. lapping areas of an IPS and DSS are only iolated elements or subcomponents of an O1S. The primary advantage of an OIS over a system in which each ofthe base olfee functions i automated Independently isthe eduction in the complesity of the interface between the office personnel and ‘the machine. Unless this interface s kept simple and is easy to learn and use an OIS may be of litte value except to the office specialise, A second benefit obtained from an O1S as compared to a WP, IPS, or DSS is the result of the fice structure the OFS is expected to automate. An office organization is designed to accommo: date office workers performing highly interactive, autonomous, and concurrent business tatks Fixed programs are inflexible, ad simplistic interactive systems are deficient in mectng the needs of the offiee worker as reqused by the dynamic office environment, General O18 Capabilities As mentioned previously, an O15 is an integrated computer system comprising facilities for pe forming office work. In particular, office automation aids office personnel in the preparation of ocuntents, information management, and decision making The types of documents to be pre pared, processed, and managed by an OIS are memos, letters, reports, and typical business forms ‘uch as requisition and purevase order form. The document, stil visually depleted in thei repular format, are stored electronically in an OIS, The methods of capturing the Information provided to the office workers produce documents with the standard headings and leteting apptoptiats 0 cach particular form The office Worker interacts with the OIS through work stations. A work sation isa program rable microcomputer equipped with office-oriented devices and software. The work station wil enable the office worker to sce and manipulate the electronic documents, communeateelecton- ically with other work stations, access data base information, and stepwise dcect the work station to perform designated tasks on a one-time basis of automatically. Such a onetime tatk might Be prepating a memo incorporating only sections of previous memo and a previously prepazed Summary report. An example of a task performed automatically by the workstation might be searching an inven tory data base to determine whether tems lated on any purchase requisition ae instock OF On joxder or whether they need to be special ordered. The work station operator covid then fl out 3 blank purchase order, using the information found on the requisition fori and in the inventory sata base. Adtional accesses to other system information could be made ab the situation ware fanted. The completed purchase order form could then be transfered to another work station Aletermined by the nature of further processing, such as managerial approval ofthe purchose In addition to eleetronic dectuieats and the traditional datatypes, O1Ss will sport veo data such as pictures and graphs. Graphic packages are currently avaiable and ean be intqrated into the OIS. Speech recognition would be a very desirable feature of an OS, since some manages are ‘opposed to Keyboard interaction with the work station and prefer dictation ad recordings In communicating with office personnel. Currently, speech recognition systems ar in avery primitive stage. of development. An’ experimental «peach understanding system, Heasay I, tecognzes ttteranees in e 1000-word voccbulary, with a corcect interpretation tate of approsimatsiy 90%! ‘Other alernatives to the Keyboard are touch serens and ight pens Communications between wotk stations can be instantaneously carried out by means of elec: loonie mail or other eletronie-type mesager sent aver telephone lines o” bY way of 2 satellite communication network. The clctronie mail system will contain known addrestes and other Pertinent information such as telephone numbers. The sending and receiving of mal is handled utomaticaly with litle interaction by the work sation operator. A “malbox" is maintained fot ach work sation, the contents of Which may be surveyed upon request. Immediate notification ofoTEMS vained noe of sing Por include S.The mated nchand be of of the tasks needs on of * pe slat vided nw it be paced aa othe OFFICE AUTOMATION 1273 newly artived mails posible if desited, and an indicator ight that shines when the mailbox is not empty is feasible. The physical configuration of an O1S could be as simple asa single processing unit supporting several work sations of a8 complicated asa distributed network of large, interconnected comput frs, each of which supports a cluster of mor stations. The degree to which each workstation can ‘operate independently of other work stations and nctwork processing. devices determines the degree to which the OIS is immune to hardware malfunctions snd downtime. Each work station will have local databases and data objects and indreet access to Information local to another work Station. The design oF a particular OIS wil determine which data bases are available to each work Station at any piven moment and how that information may be accesed and use, Multiple tasks may be pecformed in parallel at individual work stations, Tis will occur most often when the office worker describes entice tasks or parts of tasks so thatthe procedures may be handled automatically BY the work station. When the work station encounters an unanticipated discrepancy or situation, it wil notify the office worker describe the problem, and await further Tm addition to the concurtent processing at individual work stations, parallel processing tay fccur when multiple workstations. perform operations on a single transiction. For example ‘consider the purclase order that requizes the independent approval of two managers, Once the purchase order is ready for the approvals, it would rive simultaneously atthe two managerial ‘ork stations. Fach manager could then approve the purchase put i into a "wait state” for future ction, of Dock the purchase by denving approval. The ran point is thatthe two approval pro- esses do not necessarily have 10 be staged sey; instead they can be petformed concurrently. IF fone manager should disapprove the purehate, the other could be notified, and the purchase order handled accordingly Summary An O1S cat be viewed ata distributed network of programmable (intelligent), ofice-orented work Stations capable of communication with other work stations. The bulk of the Work supported by {an O15 focuses upon preparing ané managing electronic documents These documents aze set {© the work stations where work Is performed. The route of exch document Is denoted by the ct ‘of actions requied to process the document. The tasks asioclated with document processing may be autonomous and interactively performed. In addition, other faites may be integrated into an OIS to automate alfice procedures. The OIS supports tle concurrent processing of tasks assigned fo an individual work station and of independent tasks associated with the processing of docu iments for a pasticular transaction, The Citibank system discussed next is an early attempt at ‘eating sn automated O1S, Citibank’s Management Work Stations In 1976, the 12 management work stations (MWS) installed at Citbonk were intended to alleviate aifce workers burdongome problem of managing and Frocesing the paperwork nevded to handle fustomner taneactione® The MWS systom semves sun Interesting introduction to OIS prototypes for a nuinber of reasons: ust, iis ina reallife business envionment; second, it Was designed to ‘ve s busines problem and aot simply teat ts sywnpromecand thd, tinegraved several business Tunctinns into a single system withthe end user, that I, the office worker, in mind The MUS stem eliminated many ofthe ieficiencies of the traditional paperwork system by providing (1) automated internal correspondence, 2) am electronic mail stem, (3) memo pro> essing, and (4) Tinacal and other report processing. The MWS system has builvin capabilities for word pocesting, document creation, editing and iting, Business planning, distributed calendar maintenance, tris development, cost-benetit anolSss,Iianckal reporting, hardcopy production, find vhe trannigsion of dacuments and messages Ison one work sation to another. Each manage trent work sation consists of a PDP-BA procesor, tho keyourd-dispay unit, an impact printer, ain Toppy disk dives The MWSs are connected hy telephone lies and shared by a manager and The MNS functions were designed to be consistent with the manual paper-based counterparts, 0 as to mininize the adverse effects of the transition and thus increase the acceptability of the System, Citibank observed secretaries and managers tained to use the new MIVS system inorder f0 tauge the eflect of its introduction. Citibank concluded that the secreanies have showed “more Flexibility and adventurousnes in using the system Sian managers have ‘Citibank officials consider the primary advantages of the automated office over traditional ofice methods fo be its impact on control, security, expense, and productivity issues. Electronic Hes sere found to be relatively socure, and electronic forms were shateable and safe fom accidental Tose Speedy ling snd retrieval ofoifce information and the other capabilities ofthe MWS system1274 COMPUTERS AND INFORMATION PROCESSING SYSTEMS, Improved worker productivity and customer serve. Together these advantages seem to justify the However, the limitations of the MWS system are many. Reliance on basc security and control mechanisms may be overly simplistic. The electronic mal system, to be more effective, needs to Be able to handle external documents and correspondence as well as those generated internally. The ‘MWS capabilities neod to be extended to encompass the fll spectrum of business operations and {o permit the user to "progiain” the MWS to perform some tasks automaticaly. Clealy shown, however, is that an automated office, even ths rather basic model, is practical and beneficial and will help se the tren forthe OIS of tomorrow. OIS: An Evolving Trend [As demonstrated by the Citibank example, a progressive and experimental project for its time, OSs are still inthe developmental stage. Much of the technology required to make an OIS a reality is available, The degree to which the unsolved technological and prychological issues aze resolved wil greatly determine tho useTulness and acceptably of OISs. The technological isues remaining to be answered span several disciplines and are intereated. The computer scientist developments in natural office languages, communications, and hardware suitable fr the office wil strongly influence the roles and responsibilities of office workers and the physcth organization of the office, Accountants and auditors, concerned with the possiblities of Slectronie fraud, need to ensure that intenal controls ze implemented. This wil require adequate hhardware and software support aswell as managerial control of sensitive information. Research into the psychological impact of the OLS on office workers is required so thatthe office workers" capabilities and views support its successful Implementation, Sociologists may beable to provide recommendations based on thet research into man-machine relationships and thes effects fn the socal structure. The prospect of lage interconnected O1Ss will undoubtedly send reverber ions through Washington, initiating, or example, the further propagation of privacy lenslation fogulating their use. Managing and integrating these exentisl advancement, ndings, and tech nologies into a total OIS are themselves complex problems without immediate solutions “The current Limitations of O1S implementations are ast, But certainly not insurmountable, As work continues, answers to current questions will be found, and OISs wil evolve, incorporating ‘each new finding and expetience. Since most of the formative work is done in nboratories, exam ining state of the at prototypes i appropriate 12.7.3. O1S PROTOTYPES Two OI prototypes ate ditcussed in this section, One, Officetak-Zero, comes from an industrial setting, wheteas the other, SCOOP, was developed in an academic setting Officetalk-Zero Ofticetaik-Zero isa prototype O1S that was designed and implemented by Wiliam Newman, Tim Mott, and employees of the Oifice Research Group at Xerox Palo Alto Research Center (PARC). The Offcetak-Zero project began in 1976, and the prototype was in operation by June 1977. Its goals aze consistent with chore designed into Citibank’s MWS ystems. Offietalie Zero isan elec: {ronic aid t0 performing routine office tasks, silat in manner to the traditional way of pet Forming the equivalent clerical tasks. The fundamental object in Otficetak-Zero Is the electomic document, The Officetalk-Zero commands ate for the management and preparaion of documents land for their transfer between work stations. A majot portion of the Officstalk-Zere development effort was invested in constructing a man- ‘machine interface that Would erable the office worker 1o use all of the system's facilities simply tnd uniformly. These goals were achieved by developing a Wotk station supporting desktop acti ‘ties and through access to an inesrated system providing the user witha large set of office fune Configuration of Officetalk-Zero Oicetatk-Zero is configured as + network of interconnected minicomputers, each supporting a single work station. The minicomputer employed is the Xerox Alto with 128K 16-bit words of fain memory and-a 2.5 megabyte disk for external torage. The system supports both personal And system data baer for the storage of electronic forms, electronic mal, and information ‘concerning each authorized user of Oificetalk-Zero, Communications between work stations are ‘conducted by means of the electoonic mal system and the transfer of electzonic documents. The ‘work station it an advanced CRT device that electronically simulates the office workers desktop. ‘The user manipulates forms on the CRT device by employing a pointing device and a keyboardYSTEMS iy the { controt dstobe ly. The ‘ons and ‘town, wiatand ‘reality resolved related Inedware and the ites of lequate ve office able to effects roverber ‘lation nd tech able. At porating steal van, Tip PARC)? 977. ls an see ot per simply func orting a ords of personal Tesktop OFFICE AUTOMATION as Work Station Description of Officetalk-Zero The contents of the electronic desktop sre represented by the dsplay of forms on which work is currently being performed and four fl indexes of forms avalable at the work station, for exam ple, incoming mai. outgoing mail retained forms, and blank forms as shown in Lxhibit 12°71 The system automatically updates indexes whenever the work station operator isucs a command ‘causing a form fo be moved from one Me to another. Lach ent inthe index consists ofan action Feld specifying whieh Orfictalk-Zero command may be uted to process the form and pertinent descriptive information about the form, Documents and the repository lists are displayed in rectangular windows on the CRT device, To start work on a form, the user indicates the Torm and the setion to be performed by pointing the jutsor to the action field of the appropriate enty of any index. Ifthe specified action isto make a form available for procesing, then the system ill digplay the form in a new window. Each window includes a "menu" of Offcetalk-Zero commands applicable to the form, Pointing the usor to the desived function displayed on the mem causes the work sation fo enter the apote priate mode for interactive processing or prompts the work sation to execute the selected function automatically. The user can enter data on a form by first pointing the cursor to the fel that is to contain the data and then typing in the data on the keyboard. Datatype checking is automat cally pesformed by the system so that the entered information isin accordance with the forms specifiation; for example, 2 quantity field ma) contain only numerical data A provided forms caltor describes the standard letterings and format ofthe form and the style of ech field on the form. ‘Opened windows on the CRT display i forms or indoxes may be adjusted by the office worker Exhibit 12.7.1 Officetatk-Zero's Work Station Display Personal Lever | ‘Tom Smith ero | lensed Doe Reauistion Pete Peels Recipient ‘Subject [Sent [Tee [Dick Tins | “Putehose Requet | 5/00 Trace _| Sally Hawkins | Transfer Request| 7/16/80 ~ Tn brseet 76/80 _| ‘Sender Receves_ |-_____.} = z uit [Dick Tims | Putenase Order 1/18/80 Trace [Tom Smith} Employes Referral | 7/16/60 Date [aati] festive Fa Gooy Fe send — ssouergmecreey12.76 COMPUTERS AND INFORMATION PROCESSING SYSTEMS so that the form can be seen in its entirety or in part. A form may be moved around under the Window for viewing vatious pars, or the window may be relocated on the steen, Overispping windows will produce the sae effect as overlapping pages on a desktop; the ant window moved wholly visible and hides those portions of the windows beneath it. The workstation coer conte the movement of a window and the form within a window by pointing a cursor tothe approptite 1nca(s) of the screen. When finished with a document, the user may fle or mail it. een the window and eausing the document to dissppear from the screen The work station possestes another interesting eapabiity’ I allows the usr to afix signatures or freehand illustrations to a form. The freehand markings can later be removed without stering the form's original condition Capabilities and Limitations of Officetalk-Zero Ofticetatk-Zero provides the officé worker with the base functions requized to process documents sifectively. These capabilities inchide a text editor, eletvonie mail with a tase hinction tet enables the sender to monitor the location of a mailed document, data entry capabilites, and functions for copying and filing documents. The limitations of Officetalk-Zero appeat not to be the result of shortsmnings in tl n ofits existing capabilities, but rather the oi sion of other desirable capabilities. The nonprogramming environment of Officetak-Zero does ‘not allow the user the Nexibilty of defining personalized office procedures and functions to one hance the effectiveness of the work station System for Computerization of Office Processes Zisman developed SCOOP at the University of Pennsylvania as part of his doctoral research. The system streses the automation of office procedures as well as of office devices’ The heat of SCOOP is a modeldriven monitor that traces the progress of office procedures and automaticaly executes portions of them at the appropriate times. By 20 doing, the computer and the work sation operator interact in a joint undertaking. The system description for driving SCOOP's monitor comprises document definitions and office aetvitie expressed in a nonprovedual lam sage, The underlying formalism ofthe system desription isthe Petri net augmented with single and compound predicates whose tcuth initiates the exccution of comesponding ations, The Predicates ean be used to detect certain events such a6 the existence ofa fecord, # value st By a {ransaction, or the passing of a set amount of time, Like Officetlk-Zero, SCOOP has singe, uniform interface for integrating ihe specist-purpose systems into a unified system. The overall system supports electronic message passing, document manipulation, file setvces. and other functions. The creative idea of auromatially monitoring office procedutes to asst the office worker is extendable to monitoring offle work for compliance with security and internal control speciflations This isu i explored in Section 12.75. 12,74 THE OIS USER INTERFACES Yarlice this chapter stated the essential properties of an OTS: It must be flexible and powerful nough to meet the user's nevds, and yet simple enough thatthe office worker, a nonprosammen, n use it effectively. Historically. interactive application systems were developed under a systems analyst's supervision and management. Th systems analyst's task was to survey the users splice, tions to establish the necessury system capabilities and to design a system meeting those needs. The system design was a blueprin from which the computer technicians programmed the system Uhus the systems analyst served an intermediary, bridging the communication gap between the user and the programmer, If failices in communication were kept to-a “reasonable” level the system desi would indeed meet mos, i not all of the wer's needs and would be implemented as Specified. However, often the net result of these weeks or months of implementation effort would be an inflexible system of parameterized programs that were invoked through a kind of mena selection that only temporarily satisfied the user's requirements. Changes in such systems were often lfc cult and costly, Yet changes inthe way busines ie conducted are inevitable and will requte such alterations inthe existing syste, AA highevel office language that office workers can easily learn and use would substantially linprove the potential for efficent and effective OS development and maintenance, lloweres because of the complentiesinierent in office work itself and in natural language developinent ths attainment of such a language will not be easy. Nevertheless, the current state ofthe art makes it reasonable to expect the development of high-level office languases enabling the Uset to sot up and ‘maintain a wie variety of office application systems, and yet requiring only minimal Using ‘The languages will combine WP and data base management faites, enabling the set to define, Inaintain, and query data bases and to incorporate the selected and processed data base Informa:EMS the ing vrols the that and abe ‘toes the wot ‘ally rk oP Mar. ple va ral verol val as OFFICE AUTOMATION haa tiom into reports, documents, memos, and so on, Micro Data ate Systems is good example of a sophisticated data base management system for the contgol of complex data structures. [tis commercially available and operational on microcomputers,” Word processor and data base man agement capabilities will make it possible for office workers to develop software for common aifice applications suchas those in a doctors office ‘Two proposed high-level languages for ute in OSs ate presented nent, followed by a brief discus sion ofthe controversal lsues surrounding the prblers af the user Interface with automated OS The Office Procedures by Example Language The Office Procedures by Example (OBI) language isa nonproceduralanguage intended for wse in automating office and business applications * The result of a research project of IDM. OBE isan extended version of the IBM-marketed Query-by-Lsample data base management language. The OWE fanguage makes it possible for office personnel to maintain dats bases and program work ations to create tables, forms, and other documents with information azeady avalible in data bse. Vatious mathematical calculations at also posible, Oifle Procedures by Example ssa mat tins facilis fo save programs for reuse or automatic invocation whenever certain extcal events ‘occur within the operation, Communication in OBE is by means ofan electronic maling ste. “The fundamental object in OBE is the two-dimensional form, The user specifies how the dts base information is to be mapped onto the form by tiling in the form with sample data. The sample form specifications constitute an automated procedure, which, when executed, produces 4 document. This document can be edited further of can be sent to another work station inthe network [Exhibit 12.7.2 exemplifies an automated procedure written in the OBE syle. The purpose of the proceduce is to prepare alist of new employees artanged by manager. The procedure it up 10 ‘tun automatically once a month without user interaction The reports are sent automatically tothe corresponding managers through the electronic mailing system, The. procedure specification onssts of (1) references to two data bases, 2) quecs criteria, (3) 2 template for creating the desired form, (4) a tigger command, and (8) a send command Exhibit 12.7.2_ OBE Procedure for Personnel Reporting to Management DEPARTMENT DEFT [—conoitions | ‘CONDI _ ‘Y0U/80 < DT « 12/31/80 This the current listing of now employees who are Scliedued to work for you this your, 1980. | Suny Bird |1278 COMPUTERS AND INFORMATION PROCESSING SYSTEMS ‘The first data bose of the procedure description, APPLICANT, contains the name, status (= hired. R rejected, W = hired and working), Job position, hing date (implies status isan H 02 W), and hising department for each applicant referral The DEPARTMENT data base contains the ile and manager's name for each department. The undetlined character strings, NB. DI. D, and M, are example clement, The presence of each example element in the data base descaption and the body of the sample form indicates a mapping of the data base information onto the for ‘The APPLICANT and DEPARTMENT data bases are linked inthis procedure by the depaztment title field, DEPT. The department title feld Is common to both data bases. The feslt of this crossteferencing is that M is a place holder for the manager's name for each department. The character fin the status field of the APPLICANT data base fsa constant and states that o@ly applicants with a status of H are to be accessed by the procedure. The condition box places & further constaint as to which applicants are to be reported. Specifically, those applicants who have airing date in 1980 wil bea pat of the report, The automatic scheduling facilities of OBE maintain a Us of procedures that are to be run rou tinely. It automatically schedules them for execution at the appropriate times. The tagset com. ‘mand, TRI (MONTHLY), and the SEND (TRI) A to M command specify that each monthly report is to be sent sutomatically to the manager that the letter A addresses. Whenever this proce dure i executed, eportssimat to the ones shown in Exhibit 12.7.3 ate created. The exhibit also Presents the contents of the databases used to create the Forms. ‘The OBE language provides means for defining, updating, querying, and delesating user access rights Co data bases. Data ase integrity can be preserved through the use of integsty constains Potential integrity constraints include the editing of the update data for reasonableness of validity Exhibit 12.7.3 Sample Input and Output for OBE Procedure for Personnel Reporting APPLICANT | NAME wriw [Postion | pate | Derr Sandy Wikon | Secretary 03/2080 | acer burt Witiome | OR Accountant Sue Fieide H Accountant | overs: | acer | Ton Smith w Merasst oyisiso | acct foe Peele H MerAsst 9/01/80 | PAYR Geld Farrel H Accountant | 07/18/80 | ACCT DEPARTMENT fe or MANAGER ‘ACCT | Kathleen Roberts NGR | Joe Doe [_ PeterDavis Dear Kathleen Ro This is the current tsting of new employees who ate Scheduled to Work toe vo this year, 1980. Srey Bird Date Sandy Wilson Secretary 03/20/80 Gerald Faret Accountant 07/15/80, —I Dear Peter Davis This isthe current listing of new employees who are Scheduled to work for ye this year, 1980 Juney Bird Name Position Date Jos Peels Marat 09/01/80OFFICE AUTOMATION 1279 of values and for interdata and database consistency. The intesrity constraints may be established Sich thot they are unconditionally enforced or enforced for only particular data base operations As noted inthe preceding discussion, in addition to executing procedures automatically based on Specifed timing criteria, procedures and other OBE actions may also be executed automatically Whenever specified vents occu, for example, an overdrawn account. The conditions controling {he attomatie execution of OBE actions and commands may be simple or compound. The Business Definition Language ‘The Business Definition Language (BDL) developed by IHM at Its Thomas J. Watson Research Center 16-2 high-level language designed specially for business data processing” The view of busines data processing manifested in BDL centers on the fundamental observation that document processing Is the primary operation in busines data processing. Documents are prepared and trans hnited singly and in groupe between departments and divisions as a means of communication. “These same documents are later used to support recording and crosszeferencing transactions. The ‘humeral operations involved tn the preparation ofthese documents are usually very low level fn objectine in designing BDL war to create a language office workers could use to write proptams that would be structured, modularized, self docimenting, and easly maintainable. In pursuing these goals, one means employed by the designers was providing the BDL programmer with a language in which whole algorithms are waitten in meaningful problem solving terms. Even though BDL vus designed or general business date processing, the design philosophy employed ints development isa Blueprint sutable for the design of future OTS languages. "he basic objects in the BDL are documents, steps, paths, and fs. Documents serve as input and output mediums for stepe and programs. A'step is. sequence of actions for the manipulation ‘Of documents, and a program isa soquence of steps, Lach step action is either irreducible, that is, {primitive step, oe 2 composite of more primitive steps. An irreducible step represents standard, predetermined document trantformation of a userartten routine describing the input and output behavior of the step. Composite steps are reducible to primitive steps. Each step corresponds to an organizational unit such a8 a department oF work station. The nesting of the composite steps Is intended to reflect the hierarchical stuctute of the oreanization. A path isthe representation of the flow of dacuments between steps. Fes are used for long-tm storage of documents, Major Components The thre major components of a BDL system description are the form definition component {FbC), the document flow component (DFC), and the document transformation component (DTC). The FDC enables the user to define templates for creation of blank documents. Each focuments design 1 developed by drawing it on a grape device. Ties within a document are deseribed hy the type of data the field isto contain and by ts format. Groups of fields ean be defined as single enti, anda sort fatty ean be used to order the entities. Provisions also exist for describing floating fields and for handling form overlow (the extraneous information spills onto a user designed overflow form). The DFC application system is graphically defined i ts ofthe organizational units involved in the processiny and Mow of documents. The base distam constructs are rectangle, circles, and Strout, sed to represent steps ies, and pats, respecively. Torked data flows indicate that a apy of the document Isto be sent along each path Identified. Exhitit 12.74 presents a DEC Fepesentation of a sul processing system {ertica in terms of move elementary steps and daa flows. These constituent steps tay themseves be composite steps which could be forter refined in other DEC dagrams. Previously unspecii fll acceses nay be introduced in the more refined DI'C diagzams The sysem i fully described bye DIC ulen each stop is tedueTle, ora prmitive sep hi sles processing example i especially intersting n hat it combines aspects of both O18 and data procesng applications, The cistomer service sich prepares the invoice and approves edit tame’ The meee preparation step i performed by a sceetary at the Work station, The secretary Shains blank invoice and fils in the customer's nome, address, product numbers. and quantities, procedure written by the seeretary then accesce he price Mle ad enters the product unit price, thd total price ty product number, and the sales usvoie total rie. Fanaly. the secretary reviews the invoice: makes any necesary changes, and seni tothe appropriate Work sation for credit "he stelttfice. having specitied general approval policies in granting eredit terms on routine anders only affixes a specific approval to requets ease the general approval policy Bounds. For there special eases. the workstation is prosrammed to display the sales invoice in question along th de costomer’s credit story. In ether case, the approved invoees are forwarded to shiping.t7.10 COMPUTERS AND INFORMATION PROCESSING SYSTEMS Exhibit 12.74 Sales Procesing System feustowensenvice] | Tara} SH) (Surpng (PStercrrna-4 / 1 ae ! fot £9 i (Coecnemenr invenrony pe TE I ‘Shipping Order's | (raacees onbeRs t 1 Shipment Notioe's 1 RECEIVABLE I act // = 2) Shipping and accounts receivable oe data prosessingsysteme with routines difering fom the O18. procedures in that they need no autonomous procedures, but can be fixed algorithms equicing ety little human interaction. The. Accum function queues the invoices until midnight at which time the function releases the day"s entire group of invoices to the shipping data processing system, The shipping data processing svsteun fils the orders according toa feed prioetyseheme. updates inventory, an schedules shinpine, tn an automated warchouse, workers then supervise the packing Wr the otdets and notify the ystcin ae each order i filled. The system then goncrats a shipment notice document destined for accents Feeivable. Simarly, the accounts reccivable data process Ing system accepts the shipment notices as input and updates the aceountsrecevable nnd invoice fle. The DIC of the system description it used to define the actual computation of each ireducible step in the DFC. Theoretically ny programming language or mariage of propromming languages {ould be used in place af the DIC language. Thus 36 the previous example, an OIS language and o Husiness data processing lonswage could both be used to implement mined systems, (aking 3a fage of the office worker's proyrasnming sile and those of the application spectalists. The very high level DTC language developed assumes a general structure fr all steps and thus pves the pro- framing the advantage ofa builtin contol sirdeture oriented toward business data processing in exchange for the resulting programming simplllty, we accept a Loss in language flexibility The proprammer utes the resulting contol structure in defining transformations of the input ncumments 40 steate output dacuments The computations appropriate to each fet inchuded on the output document ate described using normal arithmetic nperaiors, agtepate arthmetie bperators, and logical and relations! operators applied to avaable elds on the input and output beuments. The DIC’ ran tiie mechanism automatielly readies the step for execution wien ver documents are available on alls input path, when prior system activity complete to that step.ols Hie ine le OFFICE AUTOMATION wat Office Workers and Machine Interface Issues In any opsonization, including the busines office, fornia and informal communication lines exist between individuals and organizational units. Formal comsiunications ate designed to mect certain Aaniiipated, critical information needs, with written dociavents deully serving at the medium Preparation ofthe necessary Torts is generally detailed 2 part ofthe system description. “This type of communication closely corresponds t» the documeatorented View of an OIS. The vwelldefined procedures lend theinselves to being algorithmically specified. The documents can be Structured and formatted and electronealy or mechanically manipulated, Moreover, 2 Maing stem isa natural and ficient means of tansferrisacurnents between the organiational units Oster information systems, property complete with ser-matched capabilities, are capable of et ‘seni supporting foetal eiice communictions Tie crite ise concerning formal communication in an OIS environment isthe office worker= machine interface. As formal communietions are machine representable, the question becomes wheter clerical wsers can get the necessiry support to perform office tasks with standard office Capabilies suc as WP and uscriten routines infortul covimunications-convesations, gestures, jvkes, and unwritten guidelines—serve aa Immportant organizational nevd often not adequately accounted for by the system. Dialogue is used by senior employees te provide on-thejob raining for junior employees. Stimulating conversations about business problem, suring ideas and knowledge, can give problem solvers new insights. Com ‘ersation can encourage personal involvement, concern, and a feling of group membership. These form of soc contact are esonsal to maintaining an eifective office organiation and exeniliy {information comniunications that cannot be adequately supported by a purely document oriented IS. The integration of information communications into an OIS fa major conceen in interfacing people and machines Office Worker-Machine Interface The implication of particular design philosophy and the nature ofthe cesulting OFS interface are inmportant to office personnel stalling and traits. Clescal workers and managers cannot be hpccted to postess csport programming skills, ad Yet hey heed to contol thee own office pro edures. An OIS wll sencally incorporate only specific problenvoriented functions and capabiltes ovnmeuly aeeded in an oie setting For exumplc, Qtysscy Isa kaowledgebased systcm for ad ine an individual tser with busines trip peparatons.” Knowledge of tip preparation f ull into (hyerey co that cam intellicetl esl the user Specbiied office nsstance packages such af WP, DDS, and Odyssey can provide interactive procedures to aid the ollice wotker in performing afie tasks. These procedures enable the worker to pevtorm effectively the particular tsk for which the system was designed. Users will require ile tiaining since most of the systems will have simple interfaces using basic problem-oriented commande or prompting techniques. However, they cannot cover the wide sange of autonomous flfice applications cirrently requied by husines. Depending upon the mature of the lask. office ‘norkere wl be expected to spocly and corect prostams and fo maintain thelr own automated blige procedure, This section adresse afew of the many issues concetning the design and implementation of O18 languages that ill cose the ofce workers burden in this respect. A rumber of these issues Nave been aldressed by the computing community and correspond to questions reeaeding conventional ming languages: structured progeamming, concurtent languages, progr and system doe Fntation, data base languages, and the eiaintion of tevhinkaliles such ax memory dumps fom the programming environment nmin langues are defined by thelt syntactic and semantic rules. For a programmins language to be properly interpreted by the computer, i must adhere (o.a set of rules that can be igs an signin. This need to structure and iteepeet the languages rigorously puts even limites natural language processing ouside current prowtamming technology. Clearly, the styles of conventional prowramming languages such a+ COBOL, FORTRAN, oF Pascal are inappropriate for unsophisticated O18 users, Por this reason, languages more appropriate for O1S have been described and ae reviewed here An example of 8 specia-pospose OIS language(s OBE. ts simple, mimicking style makes OBE very easy for even unsophisticated clerical personnel to master. Its well sulted to progamming work stations called upon to prepare simple forms formtted according to the OBE standards ‘The automatic task initiation featuee is abo easy to use and very practical. The problem with spociakpurpore fangsiges like OBE is that their application span is very narrow. Tasks coinciding stith the abstract mou! on which the unguage was designed can be exslly encoded: however ny Alsiation feom the langvages orginal problem formulation assumption ‘wil greatly increase the prograniing proble aot make it impossible co expres in the language. ‘At the otter end of the language contineumn ate general-purpose languages. Fora language to bei942 COMPUTERS AND INFORMATION PROCESSING SYSTEMS eneral-purpose, it must consist of fundamental operators, datatypes contol stuetures and data Structures with which problem-oriented routines ean be constructed. Translating a conceptualized algorithm into such a language is difficult; however the more genera purpose the language, the wide the range of problins i can be used to solve This benefit i obtained atthe cost of incre. ing the conceptual difficulty in the programming tsk. Matching the right OIS language to the clerical user's or application specialists needs and abiltes is eSsentil to effectively iczeusing Productivity. A mistatch wil reduge an OIS's effective ute and could cause the system &0 fal Problenvoriented BDL. is about midway between a generakpuspose and 1 special purpose La guage. Variations on BDL may provide an almost general system framework permitting the auto Mati interface of two or ore languages. As the sales processing exarnple showed, the DEC served 48 the system framework in Uepizung the reducible steps and thei tlationships Offieeoriented steps could be programmed in the O1S language, and data procesing steps sould be programmed ima more appropriate language. If each step complied with the same document input and output concept, then tie interface between the languages could be the document. On the surface this style of integrating languages is appealing because it permite the selection ol the prope language for the task and the automatic intertacing of OIS sith data processing syste The introduction of the dnclligens form could help to lessen individual tasks by embedding controls within the fon, The intelligent form would automatically perform complex edit and date consistency checks on tet and restrict the individual user's rights fo view update, o¢ add informa tion, depending upon the user and/or the form content. The intelgent form would schedule itself for work station processing according 10 specified prionties and Gurrem! work station loads and sould become persistent if neglected too long. Once the form is so descibed, office personnel would need be concerned only with primary document transformations. This would guarantee a ‘measure of consistency in processing the document and would improve data security Another noteworthy isue includes the posiblity of standardizing OIS languages to eliminate manufacturer dependency and problems involving constant, necessary traning and retaining of personnel and to increse the portability of office procedures at well ax data and procedure Informal Communications [As previously mentioned, informal communications ate necesay to performing etoup tasks inthe business office. In & puce documentoriented environment, casual memos delivered by means of an electronic mail system cannot replace soci! contact and information exchange. An O18 that could support video und audio wansmissons would heighten social interaction. Whether thi wouldbe adequate san unanswered question. Otficaofganization designs sie at increasing socal contaet ate 2 potential solution to the problem, but perhaps at the cost of some lost machine= based effectiveness. As hardware technology advances, work sation portability will Increase the Prospect ofthe office worker's home becoming the office~a station with far-taching social and business implications As first-generation O1S¢ are introduced into the business environment, there i likely tobe little Joss to the informal communication system. Implementation will probably not be complete, ad initially clerical personnel will rely on each other while they learn how to use the system effec tively. However, as OIS users become more prufiient and O1Ss more powerful snd complete the task requirements or sncil contact will decrease 2.7.8 OFFICE INFORMATION SYSTEM MODELING. Office information syscms developed around the frial lines of communication and format work Station activities are subject to precise modeling. The advantages in modeling OISS are several First, an O18 described jn a formal manner ean be checked for descriptive consistency and com pleteness. Second, a validated O1S model constitutes 3 standard and uniform documentation of the system. Third, the model can be used to further various types of analysis, For example, office elfciency specialists might use the O1S model to predict system performance, of auditors con ‘eaned with internal contto issues would be able to analyze the contro and processing structure Fourth, managers, once convinced that the OIS model is secure, adequately controlled and eff sient as modeled, wii equire asuzanecs that the implemented O18 complies withthe model. Given the complenty of 9 fulkscale office i is infeasible to perform such analyses manually. What nesded is an O1S modeling process capturing these advantages ae wel &s providing managers with the capability to restructure the organizational units and tasks under their suhority as citcur stances dictate Information Control Nets Information Contcol Nets (ICN) is a model devsloped'at Xerox PARC to describe and analyze office information Nlows.i! The ICN model defines an office asa sot of uninerpreted ectiiiesPs ata eed he sing bon ed med ‘ing iso nd vel ‘al he re he OFFICE AUTOMATION, 93 that acces and update information stored in eporioviee, Possible execution sequences of activities, te defined by “precedence constraints.” A precedence constraint isa simple rule stipulating an immediate successor activity for the Patent actly. Upon completion of an activity, one of is immediate successor activities en proceed. As with mot office data flow models, the execution of a subsequent activity need not immediately follow the teinination ofits predecessor. the work Station operator is busy or away feom the workstation, then the activity temporarily suspended The office models presented hete share a fundamental perspective on the procesing environ ment: that offices ate transaction oriented and that, even though many transactions ae performed Concurrently, the activites required to process each type of transaction can be adequately de- scribed without explicitly considering the concurrent processing of like transactions, The models ave designed to depict all preconceived processing for a particular type of teansaction, Specific Knowledge of how each type of transaction is individually procesed can be used to infer know! edge ofthe highly parallel office environment An ICN can be represented graphically or as mappines defined over a set of activites und a set of| repositories In ICN. diagrams, rectangles represent repositories; tslangles represent temmporary repositories; labeled circles denote activities; small, hollow circles denote conditional branches {hoice nodes); soln! lice connecting setivtes denote precedence constraints; broken lines denote repository accese and updates: and a small, sold circle denotes the stat (end) of parallel sctvity sequences Exhibit 12.7.5 isan example ICN graph. Inthe exhibit, activity a, begins upon the arrival ofa departmental request. Activity ay simply logs the request in a logbook. Upon completion of 4,43 Is Iniated, resulting ma typed requistion, which is then stored i the requisition form repository Activity #3 prepares « puschase order form. Choice nade ag follows ay and establish whether activity ag oF activity as isto follow. The choice depends upon whether the fequested tems are in stock. The branch associated withthe instock choice i assigned a probability of occurrence of 75. It the ouvof-stock branch i taken, activities 24,29, and ag designate that the purchase order is to be sent to another worker for verification and approval; otherwice, activity 2, completes & work order form for inventory. Note that the activites ace only vaguely defined. The designers Exhibie 12.755 Purchasing System poor easton) \ communal bob i ! | las [== ) sad ch (eee ae “|a7 COMPUTERS AND INFORMATION PROCESSING SYSTEMS Poa IGN sugRes that» vagueness is advantageous since the model can iustrat infomation flows without obscuring the basic Hows with overcoo ote fated operational detail, saat ows have developed techniques to automate the analysis and optimization of ICN. Fade ite ICN modes sre intended to be optimized according co three Kins earoe eeN {Rimations: eutomerion, reorganization. and sireamtining. Automation taniorneions weve imanwal actives with corresponding automated aces. Reorganzation amie ee ities to enhance some particular property ofthe model such sy eoncurtent eran er cam ateonmation ar ued t simplify the ICN by eliminating wis aeties aed ea communications overicad ( wubit 12°76 isa result that could be obtained by automated optimization transformations to Ee archasing system of Exhibit 127.5. Activites ag and.ay on ExnDIt IETS oe nes oe tritibl 2.7.6 within activities anda, The OIS communications system outlined eset permet Shing the prnaced ryusitonactity ey) with respect fo the los tegen acer es Hay Gealel presi of actntis au and ay, the instock check ad lrequel segs wc 12 S,aue tansformation betons to the rergatzaton category: The pinay advan’ eto tetracaine sto reduce the elapsed tne requirement completes tamacton: magi eae {evn felled forward through decision node a since the catalog needs t0 be mectond se ee imlasgetfems ar out of stock. Since activitiesa and ay neces the eaalopsand sein as Aeclate successor to activity ay, the catalog information obtained by acthith vn oe bee Maret Purchase order form inthe temporary repository p for ase atay. This Gantiornesn splits competion for aces to the catalog and Wil mate the catalog information eet abe £0 activity ag, These preceding transformations were designed to feduce hie etna en, tnd are examples of streamlining ttansformations ts ree the JEN siazram with branching probabilities and average service times, it is pos Eis caulate the expected service time to proces a single ansaclon Covedes Fanta ison Prrseaena” SAamele of sich a augmented ICN diagram. (Fle are omited for singlets y Tig Preach Surges are labeled with the branching probaities and te expeced some sen ish secied activity. For exanile, the servic time for aciity @5 16 LS mis ane a tage OF actitity ay initiating activity ay oF ay are-3 and. 7, respectively. Thus the epee es tang Sutuzed to complete actisiis ay and ag or ativiies ay and a LS 3(5)"e TLIO Ip, imilaly, we could calculate the expected service time ofthe eatie graph" and ele notation, the computation performed a Fellows. ‘expected service time ofthe graph = W[2(4!)] Exhibit 12.7.6. Purchasing "System of Exhibit 12.7.5, With Automated Optimization ‘Transformations [oxen --- +0 /\ Cor \e- = — LIND CL ee L-- + = =f Prete cr]fen shit vs sna Tas the red i sal ad OFFICE AUTOMATION r3as Exhibit 12.7.2. Augmented ICN Diageam and Expected Service Time Calculations (es) T YL ny We. 81s 5 19) vel o 0 0 oF 640 9 6 0 0 ol oo 3 3 oo oo 00 0 oo 00 0 9 6 4 0 of 0 0 42 28} WP)=(0 0 0 0 9 r= 6 4 12 28} Weih}< [2 48 6 6 28}=108 where Wy «5 = probability of entering the graph at atisity y= probability of directly Following 2, This type of analysis may be useful in analyzing oftice max ascertain the effects of introducing parallelimy into mod vidual and eroup work loads and processing times, and to ru Such elementary analysis is useful, but very limited. For example; consider arottin ake, {tansactons are proceed on a pronty bass. The transactions pionty may be bated upon setae syle, such 2s a customers credit rating, or on more complex, mulple ctitea such the hase amount. the customer's purchasing history, andr earfont inventory leacts Now’ Son eae {he prohlem by adingarval rates for seasonal or sporuiic transactions Calculating the eecered slased une to proces a particular order stansacton devs eatemely effin sad Ne ae ‘questionable. Cutenty. simulation techniques are eniployed to anaiyae such mouse OIS Simulation vel characteristics. It is possible to to answer questions concerning ind iy system throughput eharacterstis, System simulation isan ite istiplinary technique widely used to asst in measuring system per: formance. The ability to simulate presumes the ability t0 niodel the system: Indurisia sear ‘nd operation rescarchers use simulation a a tool for analyzing such systems as machin ee eee Accounting researchers use simulation techniques to study lsnsaction procenine seers ‘measure the cifcctivness of the err detection and correction fcliies comoned ee ee desi. System dsigners also rely heavily upon sinulntion in fine-tuning sees cae eee Common to tos simulation applications of this type isthe need to Mesaite tava ses ance and it components using measifments such ot lization, werk loads theseera oe response time Simulation inodels employ deser itions of system procedures, queuing strategies, and probability Aistibutions for such attibutes as arial rates, service times, system resoutee needs nd outing122746 COMPUTERS AND INFORMATION PROCESSING SYSTEMS probabilities, Since the automat to oie is system sia office signers ea also employ simulation techn fo those currently being simulated, ques 8-1 design tool. Detection of bottlenecks and underuilized work station capabilities can give the office designer crucl msights inte ere izing office and work station tasks. The statistical reslts of two simulation meedae opeierey lander the same work load provide comparative measurements for evaleaing one agate the cakes Work Station Simulators ‘Natt and Ellis propose the use ofan office environment simulator not only for observing the rele time pertosmance of an O48, but algo as a personel training foo"? The peronnel walnee ole & work station simulator that ean give the office trainee hands-on, real-time experience, Ege and eiving simulators have already proved the effectiveness of simulators for leaning papers The indvantags of traning on 3 work station simulator ee a Tolows 1. Beemts and situations can be manufactured to specifications. would repeatedly perform the same ordinary tasks, spending ver i exceptional situations. At y workstation simulator ll preconceived ‘cat according to specitied frequency sitee 2 Meszurement of office worker performance. Measurement of office worker performance is an evssision of aivantags I. By presenting office workers with denticl work Wau snd eolgeiae satatcal information as the office worker manipulate the work sation, measurements sre oe tained that are comparative between oice wotkets, between individual sions, and again pe specified standards 3. 4 practicing environment for developing automated office procedures. The work station ‘imvtator would peemit the definition and execution of office proceduies by the operator af the ‘wink station, This would enable the tinince to gain valuable programing and debupeing exper fence without incuting the costs associated with mismanaged informations. 44. A coniolled enironment flaboraiors) for studying human performance in relation to ‘honees in the work station design or capes, 5. sl ststems develapinent tool. As work dures, these procedures can be used to si cedures Using a real O1S, the teaine ile time handling important vents and situations would siation operators develop and cefine automated proce ulate unmanned work stations. In this way “rteetness and completeness canbe tested Ina real OS environment. A facility for reabtime interactive simulation at Xerox PARC. The foclty, Baekill Station or a dedicated device such as fashion. The of a distributed office system has been developed prises a network of nodes, each of whichis work inter of a data base scrver, One of more of the work ‘operated by automatic procedures that model the would-be operator n'a ealisee maining unautomated work stations serve as simulators that interact withthe entire The realism of the simulation depends upon the automated work stations" ability to reply to the simulator work stations specific demands, Since the constuction of the reply is comtinges ohne the natute of the demand, its imposible for automated procedures to reply realstiely to oven ‘mation communication events such as memos. Howevey, a certain degre of rellam eon be schee od By identifying some basic chatacteristics of the demand and answering with prestored teptig os replies taken from a database of historical meractions between teal work saticn sperntors The Backalk facility possesses many of the abilities and properties presented ia the ist of advan tages above, i a single work station simulator mode, enables the workstation persion to pee grain procedures; access data bases; manipulate forms; transfer them to other sovomered tece ations, which in tuum ean generate nes form tanserfoums co other Work Hato Weclodin, "he work sation simulators and destroy forms ‘Theoretic Internal Control Model—TICOM I An important aspect of any OIS i the suitability, accountability 1 the system. The Foregn Conmupt Practices Act of 1977 sertin intetnal conzoleateria that lessen the possibilty of 8 result, every OIS operation ie subject to enteral leg ext ‘riteria are imposed on any OIS in addition fo tie sual engi The legal restrictions on office procedures, whether automated or not, ae not the ony cater an OIS must satisfy. Fito managers also will wish to verity that only authorized inde hee ee company assets for valid purposes. When document processing is executed by sachine ramee ies ‘manually, the difficulty in verifying imernal contra procedures becomes mere comeler ig ovzoborated by increasing “computer thoi.” As corporations grow mare complex, evel, auditing tims in the traditional fashion-with involved, labosous flowchartoy wel cers Aeseriptions of document and asset flaws becomes more difficult and internal contal protections requires fiems' procedures to satisfy misallocation of corporate assets, As tra, Consequenty, significant design ectng and convenience constints oF spec part ors Tr be doe in wil pur the eps Tie Mec! The cha simu the ont ‘* 127 ind firm expr. FER thes Or really Th abou thes cus Tange ‘al have proce audiOFFICE AUTOMATION 122.47 rhetefore stistving lesal constraints and management desires will oon require more extensive computerasisted auditing. There are two important advantages to automated auditing. Fist, the ‘feck preciion, and cost of computer pracesing vompare favorably with their human counter- ‘Siti Second, inerfacing an automated auditing and internal contol system with an O1S will tin a nesely continuous audi considered to be idea! in auditing literature. In this way the ‘O18 may be modified to satisfy legal constraints and permit economical internal control vert “TICOM 116? js an example of a computerasised internal control verification system. In a sense it proceeds in mush the sane ways human auditor doesn establishing that certain internal fontol Conditions are satisfied. However, the analyte capabiities ave automated and therefore can te lnecfaved effectively with an O1S, The result ofthis conjunction is continual proof thatthe O1S Goes not violate sound management principles of internal control Intuitively, when fiom employees use the OIS to process documents, the result ofthis procesing will be examined by TICOM Il. The latter wil then consider Whether the specified processing ‘mplies violation of sound internal control. For example, i Is coasidered good practice for the purchase onder svat to the tecewing department to. contain only quanuty data. The receiving Eeparument must enter the actual quantity recelved from the vendor and not merely copy from fhe uaity listed on the purchase order when completing a receiving report. Thus, ifthe receiving Gepartment attempts to access quantity data, TICOM I! will immediately fag and prohibit the feces as a violation of internal control, This prohibition exemplifies the interaction between TICOM It and an OLS in thtating internal conta violations. Mechanies of TICOM II Modeling The TICOM 11 modeling procedure is generalization and semiautomation of the flowchart and rarative procedures auditors use (0 describe and evaluate internal contol systems. The modeling input is 2-descripton of the dutie and tasks of individuals within the firm. Instead of the flow- tart, Roweves, tis modeling i expeesed inthe TICOM Il modeling language. The procedure is Similar to Rowchartng these duties just as waiting a lower for an algorithm i similar to writing 2 compater program. An example ofthe TICOM IL modeling i iven in Exhibit 12.7.8. 1 shows the relationship of the floweharted boxes representing document processing and the commands Of the TICOM Ml modeling language. A complete description of the language and an extended sample ean be found in Baley ct ak ‘A unique aspect of te language is embedded in the WAIT expression Mlustrated in Exhibit 1257.8. The modeling proceduve currently used fos Internal control descriptions involves the simul faneous construction of 3 fowehart for the entice fim. In the TICOM If modeling procedure, iuival tasks ate described separately, and these ate integrated into a complete madel of the firm by an automated algorithm The links between individoal tasks are daerminod By the WAIT xpressions. In Exhibit 12.7.8 we see cletk 1 waiting for requisition approval. This informs the Computer that the TRANSFER of approval fits ato clesk U's ties, Thus, by matching TRANS- EER to WAIT exprossions, a model of the firs interna control system can be tecovered From {ie indvidval descriptions ‘Once the firm models pieced together, TICOM HI then analyze it for consistency and complete ness. This processing phase involves checking various items. Fst, it establishes whether each com- rant perform a vali! tsk, whether the document processing involves possible actions. For {vceaple, ian ten on «purchase order, such a “vendo.” iv altered, its established that this item teal exists on puschase order. Sevond, it determined ha every WATT expression has a cozes ponding TRANSFER that satisies it inaly, each document inthe system is examined to ec that Tes welt defined, Consistency and completeness checking i 2 large part ofthe auditors task, and thus the machine sieorthm easly gsusts the auditor with automatie model constuction and the fn testing "The major value of the TICOM Il system, however, i ts ability to solve automatically queries bout the internal controt model. This ie executed by means of query language. The query lan- fusge is expressed in'a predicate calculus Format, with ll predicate (eg. purchasing (p) means p Fin the purchasing department) and terms defined in the modsl context. The types of queries the system cam ansier mclude the folowing: (1) Can an individual or department access a set of Alocuments and assets? (2) Can e member of 2 glven set of commands follow (precede) another Sec of commands? and (3) Can «set of conditions characterized by the predicate eaeulus query fanguage ever be fave in the firm model? These queries respond to most internal control sues Curtendy examined in the lteatute, The mechanice of query answering and the computational ‘Somplesity are examined in Bailey et al® ‘Although the modeling provedure, consistency and completeness checking, and query processing have been extensively exatnined, the interface to an O1S i ll under investigation. Intuitively, the proceduve sto preprogeam the TICOM IL system fo answer internal comtol queries that concern Pisitors Then, every time the OIS ie faced with document processing, it willbe established that17.8 COMPUTERS AND IN "ORMATION PROCESSING SYSTEMS Exhibit 12.7.8. Example of ICOM I Modsting Ne A Transfer req 1 to manager fro Hy a rite bee ence this processing is consistent withthe pe Thus every document process ‘ng request may be atively shown to satty accepted accounine prace ‘ete inportant ate of TICOW I incoies tne design of the Ols Once an OIS is proposed, it apts nade in the TICOM i modeling language. Relevant internal contol ca eee ee Soaiona tte Proposed OLS. If vations ot acepted accounting princples ae eee Taegara aly reeautions can bc inplemented unt the OSs dsomed asoptatle ake es tral a ufunetions not only a a secur enforcer for an opeational system, bat soar ene, Progranimed security issu, Accounting principles form an important part of any 0 the links between electronic and mua! document lshins satisfactory internal control. Veriying that sec tm adition, ICOM It may be used av system, Finally. it maybe incerta iconal contro eiterastistaction petating management system. tn an OIS, ocessng compound the dificully of esta h contels exist is the pusposs of TICOM I csign ool, by detecting failures inthe internal contol eed with an operating OIS to provide continuous opdotnns ot rid desirable, continuous avd 12.7.6 INTRODUCING AN O15 INTO THE OFFICE SETTING ramicarsmation of an OFS is « much-liacused subject among business and office managers. ts Sete ee eteesting problems are 11) the magnitude of the automation task; (O) the aeectaee cnarainezementaton lead time: (3) the wnagntude ofthe cesel taining) the area ae syatalabl technician fr technical suppor: (8) the need to interface IS apptieninat nn apetrtlons: (6) the need to devlop audtablty criteria and techniques; and'3) ihe ee Brongi aatel and technical experince in thee areas. t would be folly for ur toseeearn Coste Procedures fr succesful OIS implementation; these wil be derive oly taney ee Shrstionce. However, careful planning should eliminate many posse pitfall ond lessee Sequences of inevitable, unforeseen vbsacies, {orf the most postive steps an office manager could take would be to formalize job tion spcifiations. & fundamental objective af business as before, but with the ad is not primarily intended to alte althoush this may result It may be free office workers to perform descrip an OIS is to permit the office worker to conduct ronie processing eapabilties. Automating office work individual tsk assignments or responsibilities sgnficanty, teue that increased productivity brought on by automaton wal ‘mote important tasks, but this wil only add to thelr curent voleg oF el 3eEa3 Pr i RISTEMS dit ed, vay Hs, tb. "iL ral of bP OFFICE AUTOMATION 129.39 There(ore accurate, precise, and complete jab descriptions are an essential input to the automation cycle Its strongly suggested thatthe job descriptions be represented in a uniform and Tora manner so that the ofice worker and application rpscalist can automate directly from the proce dura specications ‘second desirable property ofthe job description specification would be its machine acceptabi- iy, using a language similar in style to a computer progrinming language: The TICOM It begoeas Inay be particulaty useful in this respect. Such a representation would provide notemly aa ope tional statement of ihe procedures, but also asystemvige cantol of date ow felationshpe among the tasks. Compitetke programs could check the descriptions for inconsistencies and model of te office could be mechanically constructed from the procedural descriptions, Using the computer office analysts could tien apply simar techniques to rigorously svalute and measure the syneets performance on elficiency and internal conteol scales. The validated office model and the ostens and job descriptions aze the Blueprints fom which an OIS can bs implemented The cost of sucl a project will be expensive, but cort beneficial, over the implementation eye 1 would seriously adress and partly solve some ofthe aforementioned problems of Ole rope ‘mentation, particularly. problems 1, 2, 3, 5, and 6. In dition, rigorous analysis ot curest ote Procedures could 1eap immediate yains from the discovery of iellicient and inadequate interna Eontiols, A managerial task force of representatives from all relevant disciplines should be formed to pre= pate specific longterm plans for OS implementation. Susgested. task force member: inelods Gomputer and TDP technicians, personnel and behoral experts. accountants. and auditors departmental reoresentatives, and office designers. Open communications between tsk Forces 9 share ideas and identity problems and solutions would be beneficial. Throughout the planning and development stages. it must be remembered that tt QIS is not in isell a labor saving deve {iis by no means guoranteed to sole today's office problems automatically. The philosophy and design of the O18 must take into account the needs, abilities, and limitations of the orginization and its individual members. However, if the organization is ready, the automated office wil be limplemented successfully and beneficially 12.7.7 IMPLICATIONS OF o1s As the world becomes increasingly interdependent, reliance on office-type communications, infor ation processine- anid knowledge storage and reiiesal will grow The advent of conglomerate firms has requized business offices to handle information supplies and tequeys fom all markets and all parts ofthe slobe. This has strained traditionally Isbormtensive office processing to the breaking point in some industries, for example, the stock exchanges. In banking and te stock market electronic processing and office automation have proved necessary, sissy to deal wih ‘he growing volume of transactions. Other industries ave not yet been quite so stniced The effects of office automation will be stating and pervasive, As computers direc more office Procedure. offi workers will inezeasingly find their job focusing on machine interactions: As for effects on the fice job itself, office workers willbe affeted more direction from machines pave ing documents to be processed and comments on the processing. Consequentiy. he office worker will fee a comesponding decrease in the need for sllelisnce in job performace. The machine also will actually pertorm many of the simpler tasks note done by the office worker. However, oe O18 conesptspeemit job requirements to be designed to fit the worker, job entshiment fe» pores tal variable avaiable in developing systems for masimmum production, Throurhout this work. we have described and discussed OHS issues we fel will reatly affect the lives of ollie employees and the manner in which oftics busines wil be conducted, Concerto, Knowledge of these O'S design issues should pve the reudcr, the office manager, and the etiee archer isights nto the problems as well 36 the henctits of offee automation, Although se ould not report on all the aspects of OLSs in this skort space. we have surveyed the breadth of row amd eseting el REFERENCES L. M.D. ZISMAN, “Office Automation: Revolution or solution.” Sloan Management Review, Vol 19, No.3 Spring 1978). pp. 1-16. 2. LD. ERMAN. fT REDRICK, V, R. LESSER, snd D. R, REDDY, “The Hearsuy-1 Speech Understanding’ Srstem: Integrating Knowledge’ to Recalve Uncertainty." ACM Computing Siavevs, Vel. 12, No.2 1980), pp 213-255, R. B. WHITE, “A Prototype forthe Automated Otis.” Datamation, Apsil 1977, pp. 83-90. WHITE, "A Prototype for the Automated Office.” p. 85. © A. ELLIS and G. 1. NUTT, “Office faformation Systems and Computer Scicnce,” ACM Computing Surveys, VoL 12, No.1 March 1980}. pp. 27-60.123.20 COMPUTERS AND INFORMATION PROCESSING SYSTEMS 6. M.D. ZISMAN, “Reprctentation, Specification and Automation of Office Procedures,” Unpublished doctoral disertation, Wharton School, University of Pennsylvania, Philadelphia, 1997 7. MICRO DATA BASE SYSTEMS, INC, User? Manual (MDBS.DMS and MDBS.DDL), Author, Lafayette IN, 1979, 8. M. M. ZLOOF, A Language for Office and Business Automation, IBM Research Report #35086, 1BM, Cambridge. MA, January 1980. 9. M. HAMMER, W. G, HOWE, V. J. KRUSKAL, and 1. WLADAWSKI, “A Very High Level Programming Language for Data Procesing Applications,” Communications of ACM, Vol. 20, No. II (November 1977), pp. 832-840. 10. RE. FIKES, “Odyssey: A Knowledge-Based Personal Assistant," Artificial Intelligence, in press. 11, CA. ELLIS, “Information Conttol Nets: A Mathematical Model of Office Information Fiow," ACM Proceedings of the Conference on Simulation, Modeling ond Measurement of Computer Systeme, Boulder, CO, August 1979. 12, CA. ELLIS and P, NORRIS, “Office Streamlining," IRIA, Rocquencourt, France, Novem ber 6-9, 1979. 13. G. J. NUTT and C. A, ELLIS, “Backtalk: An Office Environment Simulator,” Proceedings Of the 1979 International Conference Communications, Vol. 2, June 1979. pp. 22.3.1~22.3 14, DELOITTE, HASKINS, and SELLS COMPANY, “Internal Accounting Control: An Overview of the DHaS Study and Evaluation Techniques," DA&S, 1979. 15. J. L CASH, A. D. BAILEY, and A. B. WHINSTON, “A Survey of Techniques for Auditing EDP-Based ‘Accounting Information Systems,” The Accounting Review, October 1977, pp. 813-832 A. D. BAILEY, J. H. GERLACH, R. P. MCAFEE, and A. B, WHINSTON. “TICOM Il-The Taternal Control Langage,” in A. R, AbdelXhalik, Ed, Proceedings of the Florida Sym posium on Internal Controls, Match 1981, in pres 17, A.D. BAILEY. J. il. GERLACH, R, PL MCATEE, and A. B. WHINSTON, “Internal Con- trols in the Office of the Futore," FEE Computer, bay 1981, pp. 59-70. 18. A.D. BAILEY, J. H. GERLACH, RP. MCAFEE, and A. B. WHINSTON, “TICOM I: “Theory,” working paper, Krannert Graduate School of Management, Purdue Univeristy, West Lafayette, IN, 1980. BIBLIOGRAPHY KLING, RW, “Social Analyses of Computing: Theoretical Perspective in Recent Empivical Reseich,” AGM Computing Surveys, Val. 12, No.1 (March 1980), pp. 61-110 WOHL. A. D. “A Review of Ofice Automation,” Detamarion, February 1980, pp 117-118. WOHL, A.D. "Replacing the Pad and the PeneU," Dazamarion, June 1980, pp. 169-176