at SciVerse ScienceDirect

Applied Ergonomics 45 (2014) 26e32

Contents lists available

Applied Ergonomics

journal homepage: www.elsevier .com/locate/apergo

Design of systems for productivity and well being

Kasper Edwards, Per Langaa Jensen*

DTU Management Engineering, Building 424, Technical University of Denmark, 2800 Lyngby, Denmark

a r t i c l e i n f o

Article history:Received 25 February 2013Accepted 21 March 2013

Keywords:Production system designComplexity managementStakeholder perspective

* Corresponding author. Tel.: þ45 4525 6031, þ45E-mail addresses: kaed@dtu.dk (K. Edwards), plaj@

1 The description of this case is based on an interviwith Professor Jesper Larsen. The case has not been pbut the algorithm developed is described in Rasmuss

0003-6870/$ e see front matter � 2013 Published byhttp://dx.doi.org/10.1016/j.apergo.2013.03.022

a b s t r a c t

It has always been an ambition within the ergonomic profession to ensure that design or redesign ofproduction systems consider both productivity and employee well being, but there are many approachesto how to achieve this. This paper identifies the basic issues to be addressed in light of some researchactivities at DTU, especially by persons responsible for facilitating design processes. Four main issuesmust be addressed: (1) determining the limits and scope of the system to be designed; (2) identifyingstakeholders related to the system and their role in the system design; (3) handling the process’ differenttypes of knowledge; and (4) emphasizing that performance management systems, key performanceindicators (KPIs), and leadership are also part of the system design and must be given attention. With theexamples presented, we argue that knowledge does exist to help system design facilitators address thesebasic issues.

� 2013 Published by Elsevier Ltd.

1. Introduction

How can we ensure that the design and redesign of technology,production and service systems consider the human characteristicsof the people who use and operate these systems? For years, thisquestion has preoccupied ergonomists and human factor special-ists who aim tomove from fixing systems to designing systems. Theunderlying arguments to support such a shift in focus are that (1) inthe early design phase, there are more options for alternative workconfigurations that can be considered; (2) the costs are only mar-ginal, since changes in plans are minor compared to changes in thephysical manifestation of the plans; and (3) such a shift would havea significant positive effect on the overall effectiveness of thesystem.

The traditional answer to the question raised above assumesthat system design is fundamentally guided by specifications in theform of a set of criteria that the final system has to meet. Thus, thestrategy is to formulate criteria for human factors and add these tothe general set of criteria guiding the design process (Singleton,1974). This line of thinking, which is pursued in many handbooks,recommends ergonomic criteria to be used in the context of systemdesign. The impact of this strategy on production system designpractice appears to be limited, judging from anecdotal evidence

51501947 (mobile).dtu.dk (P.L. Jensen).ew conducted by the authorsublished in the present form,en et al. (2012).

Elsevier Ltd.

regarding facilities and systems, which continue to be designedwith limited consideration for the people who work or interactwith these systems.

Therefore, the discussion continues with regard to how tointegrate ergonomic and human factors in the design of productionand service systems. One position argues that ergonomists mustchange their focus from guidebooks and legislative requirementstowards enterprise strategies (Jensen, 2001; Dul and Neumann,2005). Consequently, issues related to working conditions mustbe formulated within the discourse of company strategy, and theactivities decided upon must relate to both the formal and theemergent strategic activities in the enterprise (Mintzberg, 1998).This also implies developing the role of ergonomist or human fac-tors specialist from an actor who primarily delivers authoritativeknowledge about personemachine interfaces into a politically re-flexive actor who becomes involved in development of the enter-prise (Broberg and Hermund, 2004). Recently, within theergonomic field, the notion of ‘Participatory Design’ has been pro-moted as an approach to secure optimization of both the economicand ergonomic aspects of work (Vink et al., 2008; Broberg, 2010).

This paper presents a system-based approach to the design andredesign of production systems to promote productivity and wellbeing, based on participatory design. The concept of ‘productionsystem’ used here is an umbrella term for all purposeful systemsdesigned to transform inputs into outputs that fulfil society’s needs.Consequently, the concept covers not only industrial productionsystems but also service systems and health care systems, as well asproduction systems working with immaterial inputs and outputs,such as consultancy, teaching and research.

K. Edwards, P.L. Jensen / Applied Ergonomics 45 (2014) 26e32 27

We have observed that it is common, when designing or rede-signing a production system for productivity and well being, toestablish a forum involving important interest groups. This forumdiscusses, evaluates and decides how to design or redesign theproduction process in order to ensure production systems thatfunction well from a sustainable perspective, i.e. economic, envi-ronmental and social sustainability. Many different tools and pro-cedures for facilitating activities in these forums have beendeveloped and tested (e.g. Beyer and Holtzblatt, 1998; Kensing andBlomberg, 1998). But there is more to designing systems than justprocedures to follow and tools to apply. The basic circumstances forthe design process have to be taken into consideration in order topredict potential risks and hopefully deal with them in time. Thispaper presents a fundamental conceptual framework to be used bysystems designers when initiating a design or redesign process.

2. Definition of a system

Within the present framework, we define a system basically as atransformation process, which transforms input to output for thebenefit of society as a whole. Hence, we apply a teleological un-derstanding of systems and an approach to system thinking that ischaracterized by Jackson (2000) as “system thinking for problem-solving”. The output can be material (a product) or immaterial(delivery of a service). The inputs are typically a combination ofmaterial and non-material (knowledge) objects. The trans-formation is accomplished through a joint effort by many differententities. The definition and conceptualization of entities are derivedfrom the problems to be addressed. Technology, facilities, formaland informal organizations (structures, procedures and processes),workers (qualifications, competencies, attitudes and values), and(layers of) managers can be mentioned as examples of entitiestypically used in such a problem-solving process.

3. Basic conditions in a systems design approach

In designing or redesigning a system, some basic conditionshave to be given special attention. These concern:

(1) Boundaries and scope of the system e a narrow definition willhamper redesign.

(2) Participants in the design and redesign process e inclusion ofstakeholders facilitates the process as compared to a share-holder approach.

(3) The character of knowledgee attention is given to the differenttypes of knowledge related to the design process, avoiding thetemptation of reductionism and simplicity.

(4) Performance management, leadership and key performanceindicators (KPIs) are also important design issues to beconsidered when the aim is to improve both productivity andwell being.

These basic conditions may not be formulated explicitly whenspecifying the design task, but the people responsible for facili-tating the design activities make decisions on these issues e eitherdeliberately or unconsciously. This paper is based on the assump-tion that deliberate decisions are to be preferred, as they allowdiscussion and mutual clarification of these basic conditions.

4. Determining the system: boundaries and scope

When designing a system, the designer decides on the bound-aries of the system to be designed. In determining the boundaries,the designer decides on the environment for the system, which isthus not included in the design process. These decisions also open

for identification of the entities comprising the system: some ofthese entities are perceived as circumstances not open to redesign;some are seen as outcomes e i.e. entities closely related to thesystem’s key performance indicators; and some are seen as entitiesto be manipulated in the design process.

Westgaard and Winkel (2011) give an illustrative example ofthis basic decision in relation to understanding occupationalmusculoskeletal and mental health. Based on a systematic litera-ture study, they find:

“Most ergonomic intervention studies are designed to observethe effects of reduction in relevant risk factors impacting theindividual worker, while this literature typically ignores thepotential health consequences of measures to improvecompetitiveness and productivity” (ibid. p. 262).

This implies, first, that most designers of intervention studies(which can be seen as an activity focused on redesign of a worksystem) define their redesign of the system as comprising twoentities: ‘the individual’ and ‘relevant risk factors’. In defining theserelevant risk factors, some aspects of the system are defined asbeing manipulated in the intervention (for example, weight andshape of burdens and additional tools to use when lifting), whileother aspects might be left out (for example, the individual’sphysical strength) and specific techniques developed to reduce thetask’s burdens. The last part of the quotation opens for an alter-native definition of the system. Here, the individual is seen as a partof the production system, together with the means for ensuringproductivity. This implies that the central entities now shift to suchissues as the technological level in the production flow, the workdesign, and the formal and informal systems of incentives appliedin order to secure the system’s output.

An important factor in determining the boundaries of the sys-tem is the scope of the system to be designed. Scope refers to thesize, number and variety of activities in the system. The scope of thesystem determines to what extent it is possible to change thesystem radically. A small system consisting of a machine forstamping holes in a plate only allows simple improvements inproductivity and well being, just because there are only so manyways a stamping machine can be redesigned. In contrast, a com-plete process inwhich stamping holes is only a small part allows forsignificant redesign, e.g. merging hole stamping with a relatedactivity that improves flow and enriches the job.

The designer must be aware of the scope and boundaries of thesystem so as not to focus on a system that is too narrow in scope toallow redesign, thus leading to sub-optimization. The system mustbe continually reconsidered and consequently renegotiated as thedesign process evolves. Essentially, the designer must be ready notjust to embrace redesign of the product but also of additionalservices.

It is evident that the definition of the boundaries of the systemto be designed has a key role in determining the basic configurationof the system, seen from the designer’s point of view. This is furthersupported by identifying the stakeholders that are relevant for thesystem’s performance and thereby for the system design.

5. The stakeholders and their role

According to the Oxford Dictionary, a stakeholder is “a personwith an interest or concern in something, especially business”.Within business studies, the distinction between a ‘shareholder’and a ‘stakeholder’ approach is often debated (Sundaram andInkpen, 2004). A shareholder approach focuses on one group ofstakeholders, the shareholders, and the goal is to maximizeshareholder value. This often implies that the short-term economicresult is the main criteria in decision making. A stakeholder

Table 1Stakeholder roles (Dul et al., 2012).

Role Characteristics Typical examples

System actors Part of the system anddirectly or indirectly affectedby its design and directly orindirectly affect its per-formance.

Employees, operators,1st line managers,product/service users

System designers Contribute to the design ofthe system based on theirspecific professionalbackground

System designers,engineers, psychologists

System decisionmakers

Make the central decisionson the requirements forsystem design, authorizethe purchase of technicalelements in the system,its implementation and its use

Managers

System influencers Have general public interestin work systems and productservice system design

Media, governmentand regulators,standardizationorganisations, citizens

K. Edwards, P.L. Jensen / Applied Ergonomics 45 (2014) 26e3228

approach focuses on development of the balance between thecontribution made and the return delivered to the many groupsinterested in the system’s performance, and thus on the decisionsmade that affect performance (Beevis, 2003).

It is evident that a stakeholder approach has the best potentialfor living up to the ambition of including ergonomics and humanfactor issues in design decisions. However, a shareholder approachdoes not necessarily exclude ergonomic and human factors issuesfrom decision making. In this situation, the human factor specialist,in the discussion of investments in the design, is required to pre-sent an economic argument for investing in the design, that takesinto consideration the (often very short) payback time. Henrick(2008) argues that integration of ergonomics in many cases givesa direct cost benefit with relatively short payback periods, and thatincluding these considerations only incurs minor extra cost. Severalscholars in the field have demonstrated the validity of this kind ofargument (Beevis, 2003; Alexander, 1994).

It is also argued that establishing a dichotomy between stake-holder and shareholder perspectivesmight bemisleading (Freemanet al., 2004; Sundaram & Inkpen, 2004), the main argument beingthat “a fair rate of return for its shareholders is the sine qua non forcompanies ... to fund investments required to make the world abetter place for all its stakeholders” (Sundaram & Inkpen, 2004, p.371). Consequently, the shareholder perspective corresponds to thewishes of many enterprises, not only to optimize short-term profitsbut also to show corporate social responsibility in order to ensuregeneral legitimacy and thereby continuing possibilities to produceprofit.

The stakeholder perspective requires that the major stake-holders can be identified. The focus is to identify persons or groupsof persons who on the one hand deliver resources to the enter-prise and on the other receive rewards from the enterprise. Thisthen forms the platform for analysing how the balance betweencontributions and rewards develops from introducing a newdesign.

Two groups of stakeholders can easily be identified: owners andemployees. They exchange work efforts and salary as well as pos-sibilities for personal development through their work and socialnetwork with colleagues. This identification of stakeholders is alltoo simplified for conducting a stakeholder analysis in order toidentify supporters and potential opponents of a system design. Toidentify stakeholders and the balance between contributions andrewards, the stakeholders have to be further described.

In redesigning the scheduling process for a Danish municipal-ity’s home care, it was very difficult for the operation researcher1 topersuade all the stakeholders to adopt the new process. The col-leagues had developed a brilliant software solution to reschedulehome care. It provided an optimal solution that minimized traveltime and ensured that the right care providers were dispatched tothe right elderly people, fulfilling a set of additional criteria. As astakeholder, the municipality was interested in the system, becausein one case it immediately reduced planning costs by 75% and traveltime by 15%, leaving more resources for home care. However, twoother stakeholder groups, the planners and home care employees,did not accept the new scheduling process. The planners, who hadpreviously been in charge of the scheduling process, had no interestin the system, as it would cause a reorganisation of jobs, reducetheir attractive planning job, and re-allocate them into jobs inhome care. The home care employees were generally sceptical to-wards the system and began to question the wisdom of the workplans. Normally, the employees’ critical comments were addressedby the planners, who explained the different aspects that weretaken into consideration in making the plans for the staff. The newsystem did not offer the same form of transparency; consequently,the arguments behind the plans could not be addressed. This

illustrates that a system designer cannot simply rely on the ‘hard’calculated facts but must include ‘soft’ aspects linked to feelingsand experience.

In line with Dul et al. (2012), different stakeholders havedifferent roles. Dul operates with four main stakeholder groups insystem design (see Table 1).

The different roles are not exclusive. It is possible for a givenactor or group of actors to be system actors as well as system de-signers, as the case in many participatory design processes.

A system is not constructed once and for all. Carayon (2006)argues that many people no longer work for a single organisationor only with people from their own organisation. She stresses thatsystem constructor’s work across organisational, geographical,cultural and temporal boundaries. This implies that system designis a continuous process, first, because all parts of a system are notspecified and constructed at the same time within a commonorganisational, geographical and cultural setting. Therefore designdecisions are spread over time in a process that can be described as‘sequential attention to goals’. This may not be because the goalsare conflicting, but more often because of the complexity of thedesign process in combination with the distribution of authoritywithin and among organisations. Consequently, decisions on thephysical structure of the transformations process are taken inde-pendently of decisions regarding organisation and management ofthe process. Attention must therefore be paid to how one set ofdecisions influences and delimits other decisions.

Secondly, design of a systemmust be understood as a process ofconstant redesign, where actors involved reconstruct and modifythe system in the course of their daily activities. This occurs on thebasis of the learning processes derived from handling the manydeviations from ‘the normal’ procedure assumed by the designers(Weick and Quinn, 1999). This second modification implies thatover time system actors inevitably also play the role of system re-designers.

These aspects of system design can be illustrated by an examplefrom a study of outpatient units in amajor Danish hospital (Brobergand Edwards, 2012). These units are not constructed once and forall. Outpatient units include patients with a range of illnesses andthe health care professionals who treat these illnesses. However, aconstant state of development exists in such a system as patientschange, illnesses change, and the knowledge and skills of the healthcare professionals develop. Essentially, over time, the system growsin scope in response to patients’ needs and as new treatment op-tions arise. But viewed from a productivity perspective, this system

K. Edwards, P.L. Jensen / Applied Ergonomics 45 (2014) 26e32 29

is becoming less and less effective, since the same processes areused to treat new illnesses. At the gynecology ward, an analysisshowed that treating a standard patient required 71 activities.These could be reduced to 17 activities by reducing a significantnumber of administrative tasks. Over time, the process for thestandard patient had developed to encompass treatment ofincreased scope without redesigning the process. From a treatmentperspective, it is important for a dynamic and knowledge-intensiveenvironment such as an outpatient unit to have the flexibility toimprove and expand processes to accommodate changes in patientmix. It is of equal importance, however, to take one step back atregular intervals in order to analyse the processes. The aboveexample from the gynecology ward illustrates an extreme situationwhere processes had grown organically. Such a system had gainedflexibility but lost efficiency.

Weick and Quinn (1999) address the question of how to handlesuch a situation by turning Lewin’s (2000) recommendation upsidedown. Lewin presented a model for organisational change as aprocess to unfreeze present practice, change the practice, and thenrefreeze the new practice. Based on an organisational under-standing that emphasizes the constant flux caused by variations inperforming transformation processes, Weick and Quinn suggestthat it is first necessary to freeze the present situation in order to beable to analyse it, then introduce changes, and finally unfreeze theprocess to allow it to develop until a new opportunity to examinethe situation evolves.

6. The character of knowledge

In system design and redesign, the many issues to be addressedare characterized by different levels of complexity. To handle thissituation is a challenge for the person responsible for the systemdesign process.

Snowden and Boone (2007) propose a framework model, theCynefin framework, which focuses on recognition of different typesof cause and effect relationships (see Table 2). Often participants ina system design process can have different understandings of therelationship between cause and effect. This might be caused as wellby different perception of the system as by local political issues. Theaim of the framework is to support development of a commonunderstanding of the complexity of a given challenge and subse-quently how to act. The framework comprises five contexts, whichare characterized by the different type of causeeeffect relationsrecognized by the participants.

Table 2The Cynefin framework e focussing on recognition of different types of cause and effect

Context Name Understanding of relationbetween cause and effect

Approach toand redesig

Ordered Simple order (know) Repeatable, perceivable,predictable

Sense e CatRespond

Complicated order(knowable)

Separated over time and space Sense e

Analyse e

Respond

Unordered Complex Coherent in retrospect,do not repeat

Probe e

Sense e

RespondChaotic Not perceivable Act e

Sense e

RespondDisordered Not knowing or agreeing

on what the situation isProblems interpreted frompreference for actionLatent conflicts

6.1. The ordered domain

Within the ordered domain, two contexts are identified: simpleorder and complicated order. Simple order is characterized bystability and clear causeeeffect relationships that are easilydetected by all involved as they are revealed, because they are oftenrepeated, are easy to perceive, and with experience, they are pre-dictable. The right answer is self-evident and undisputed. It is thedomain of ‘what we all know’, because all share the same under-standing of the field. This type of challenge requires straightfor-ward management. People involved sense the situation, categorizeit within a well-known set of categories, and respond based onestablished practices. These practices are stored in Best Practicesand Standard Operating Procedures (SOPs). Within ergonomics andhuman factors, many problems related to the physical design ofworkplaces fall within the category of simple order. Building onknowledge from handbooks, the recommended design of equip-ment is established. From the point of view of a design team, suchchallenges can be delegated to experts, but a participative approachmight be beneficial, first because it can inform the expert on dailyworking practices that deviate from the norm; and second, becauseit would give the people involved a feeling of participation.

Complicated order is characterized by challenges that maycontain multiple right answers, for one thing because the causeeeffect relationship is not well known to the participants eventhough it exists. The teams address challenges where it is knownthat causeeeffect relationships exist, but where resource and timeconstraints do not allow deeper studies to identify the relationship.In such cases, the team must sense the situation, analyse the im-pressions received, and respond on this basis. Normally, these typesof challenges are left to experts. The risk of such an approach is thatingrained thinking dominates the discussion and can lead to situ-ations where the experts neglect innovative suggestions from non-experts, or where a group of experts is unable to agree on anyanswer, since each is ‘caught’ in ingrained thinking or his or herown self-perception of their expertise. Instead, another approachcan be used, where the team experiments and develops a morecomprehensive understanding of the situation through games andworkshops, thus opening for new initiatives and ideas. Withinhuman factors and ergonomics, the paper by Westgaard andWinkel (2011) actually highlights two types of ingrained thinkingoften found within the field: a micro-, individual-orientedperspective versus a macro-, organisational perspective on physi-ological and mental problems in work.

relations (Snowden and Boone, 2007).

designn

Methods Central actors in designprocesses

egorize e Best practiceSOPProcess re-engineering

System experts (participation ofsystem actors)

Analytical/reductionistScenario planningSystem thinkingGood practices

System experts and system actors

Pattern managementPerspective filtersComplex adaptive systems

System experts, system actorsand selected system influencers

Stability focused interventionEnactment toolsCrisis management

K. Edwards, P.L. Jensen / Applied Ergonomics 45 (2014) 26e3230

Systems that can be characterized by simple order, however, donot allow radical design or redesign due to ingrained thinking orstrong inherent logic, which dictate standard solutions to problems.For instance, the identification of problems caused by heavy liftingoften triggers the introduction of mechanical lifting aids. The scopeof simple order systems is too narrow to allow radical design/redesign, since narrow scope induces a clear understanding ofcause and effect relationships, even if the system is ordered in acomplicated way. Such a clear understanding is inherently difficultto challenge, because the logic of the system is self-evident. Thesystem designers can challenge the self-evident logic by changingthe boundaries of the system. If new tasks are introduced into thesystem, the scope of the system is expanded, and it becomespossible to rethink the system because the ingrained and inherentlogic is challenged. This might imply expanding the system toinclude the whole set of activities of which heavy lifting is just apart. This might allow for a more radical redesign that couldeliminate the lifting task (e.g. introducing automation in some ac-tivities). The consequence of doing so, however, is that the systemunderstanding is moved from the ordered domain to the unorderedand disordered domains.

6.2. The unordered domain

Within the unordered domain, two contexts are also defined:complex order and chaotic order. Complex order is characterized,on the one hand, by the ability to explain past performance; but onthe other hand, by the inability to come up with the right solutionto a new situation. This is because the many interlinked sub-systems in the organizational context are co-producers of results.The mechanisms are difficult to identify, and the sub-systems arecharacterized by a certain unpredictability and flux. Things can beunderstood in retrospect, but the past is not repeated; therefore, itis difficult to apply former understandings. Instead, the team mustprobe, sense by gathering experiences, and then respond e forexample, by elucidating promising initiatives (Weick and Quinn,1999). In human factors and ergonomics, this might be the bestway forward for organisational development that tries to combine apositive development in productivity on the one hand, and adevelopment in job design and positive work experiences on theother. The knowledge and education of the team determine theextent to which the team is able to probe, sense and respond inorder to analyse and understand the system.

One study (Nielsen and Edwards, 2010) demonstrates thathealth care professionals (e.g. nurses) very seldom have training inoperations management and organisation. They probe, sense andrespond according to their mental model. This mental model per-ceives care as discrete care activity tailored to the individual pa-tient’s needs, but it is unable to perceive a ward as a system ofinterrelated activities, e.g. processes that can be organized tomaximize overall care delivery. This can be introduced into theteam by the system designers, with the ambition of identifyingoptions for more radical redesign of tasks and activities.

The chaotic order is characterized by the causeeeffect rela-tionship being impossible to determine, because it shifts constantlydue to general turbulence and no identifiable, manageable pat-terns. Here, the team must first act to try to establish order, i.e.reduce insecurity; then, sense areas of stability and instability; andthen, respond by transforming the situation from chaos tocomplexity. Here, emergent patterns might identify new opportu-nities and reduce future crises. This might also open for major in-novations in procedures and business models. This is not a familiarsetting for ergonomists and human factor specialists, but more sofor people working in Production Management and Human Re-lations departments. Theymay have experience handling turbulent

periods characterized by general organisational turmoil caused bytakeovers, new management and major restructuring, where theytry to uphold productivity and maintain motivation and well being.

6.3. The disordered domain

Finally, the framework has the disordered domain. Here, thepeople involved do not agree onwhere the situation is in relation tothe four former domains; therefore, the challenges will be inter-preted and addressed by the type of action preferred by thedifferent actors involved. A person attracted to simple order willargue for the need for rules and procedures. He or she is oftensupported by a person attracted to the complicated order, who willpromise rules and procedures, if they can obtain funding for furtheranalysis of the situation. A person attracted to the complex ordermight argue for workshops with participation by many stake-holders in order to reach an agreement on how to understand thesituation and then form a plan of action. Finally, a person attractedto the chaotic order might argue for experiments and their non-bureaucratic assessment in order to identify patterns of actions tofollow. Due to the different interpretations of the system, a highpotential exists for conflicts among these different actors.

The difference between the unordered and disordered domainlies in the eye of the beholder. Being presented with constantchanges and new situations, for which no prior experience exists,places a situation in the disordered domain. Over time, peopleexposed to such a chaotic systemmay develop an understanding ofthe system that might relegate it to the unordered domain. Again,the participants’ capabilities are the key to moving domain as amatching mental model allows the process of probe-sense-response to begin. The limited capabilities of the participants maybe based on lack of experience, but might also be caused by insti-tutionalized traditions and norms. Traditions and norms pose agreater challenge to system designers than changing context.

6.4. A system designer’s perspective on the framework

From a system designer’s perspective, the purpose of theframework developed by Snowden is to frame the discussion in thefirst four categories. This will allow all actors to reach, first, acommon understanding of the situation, and second, find out whattype of actions to initiate and whom to involve in order to gaininsight into the situation. A systems designer must also be aware ofthe importance of experience, knowledge and mental models. Heor she must be prepared to introduce a variety of perspectives thatcan allow the team to probe, sense and act, and thereby to design orredesign the system. The framework allows the person responsiblefor the system design process to navigate the contexts, dependingon the situation in the team and the options for reaching a commonunderstanding. The context is manipulated by changing the scopeof the system and introducing new perspectives and methods inthe team. The purpose is not to develop a mathematically optimalsolution to productivity and well being through design or redesignof the system, but to arrive at a common understanding of thepurpose, function and challenges to be addressed in the designprocess. This common understanding will then serve as a platformfor development for both productivity and well being.

7. Performance management, leadership and KPIs: don’tleave it to the managers! include people in the design/redesign

Performance management, leadership and Key PerformanceIndicators (KPIs) are important elements in organisations, becausetogether they form the management system that directly

K. Edwards, P.L. Jensen / Applied Ergonomics 45 (2014) 26e32 31

influences productivity and well being. Performance managementsystems have gained much attention since Norton and Kaplan(1996) seminal paper on balanced scorecard systems, and perfor-mance management systems are recognized as an important in-strument by both academics and practitioners (Neely et al., 2005).

Normally, these elements are not included in the system designprocess, even though they play an important role in determiningboth productivity and well being. The balance theory (Smith andCarayon and Smith, 2000; Carayon, 2009) points out the impor-tance of giving priority not only to the design of facilities but also tothe organisational design, emphasizing the alignment of compo-nents in the organisational design, also from a productivity andwell being perspective.

Performance management systems focus the outcome of anorganisation bymeasuring, analysing and acting on a select numberof KPIs, whichmaybe outcome- or process-oriented. A performancemanagement system thus has an understanding of the relationshipbetween organisational performance and individual performance.Measuring and following up on individual performance shouldmaximize the overall result in the organisation. In a sense, theperformance management system and KPIs represent the localtheory on what kinds of behaviour lead to good performance.

It is not always the case, however, that this theory matches re-ality. In a present study of a large financial corporation, we foundthat a central KPI was the number of financial products sold. From amanagerial standpoint, this is a reasonable KPI, as a simple analysisshowed a correlation between the number of sold products andprofit. After implementing this KPI, employees began to optimizetheir behaviour to satisfy this KPI by selling a number of cheapproducts, such as dog insurance and salary accounts as part of realestate loans. The cheaper products did not cover their own costs,however, and were a net loss for the corporation.

Performance management systems have a direct influence onemployees’ behaviour, which can lead them to seek to maximizetheir personal situation at the expense of the organization. So in asense, “You get what you measure, but do you want what you get?”(Edwards, 2010). Performance systems are often coupled withbonus systems, which create an even stronger incentive to satisfythe KPIs.

Performance management systems also have a direct influenceon first linemanagement behaviour. This influence is manifested bytwo mechanisms: 1) standard operating procedures, and 2)induced practice. Standard operating procedures (SOPs) are theexpectedmanagerial response to an employeewho performs eitherbetter or worse than the KPI target. Worse performance often re-sults in escalating managerial influence, starting with an informalmeeting, progressing to a request for improved performance, andending with the employee being fired. Induced practice is an in-direct result of the structure of the performance managementsystem.

Therefore, it is useful to adopt a distinction between leadingKPIs and lagging KPIs. Leading KPIs provide an indication of per-formance to come, while lagging KPIs measure the result of aprocess. Leading KPIs allow management to take corrective action,if the process is not performing as expected. Lagging KPIs do notallow for corrective action but can of course be used to change theprocess towards better future performance.

Lagging KPIs induce a certain type of behaviour in first-linemanagers when performance is below expectations, and they areforced to instruct employees to improve. Since the employee re-ceives these instructions after the results have been calculated,there is a significant time gap between employee action andfeedback. As the manager only knows the result of the process (theKPI), he is not able to offer the employee concrete advice on how toimprove performance. This therefore makes the employee feel to

blame for not performing, frustrated and unable to improve(Edwards and Møller, 2010). In a study of a large financial institu-tion, we found that sales staff members were likely to lose self-confidence when they were blamed for not performing and givenno guidelines for how to improve.

Leading KPIs, on the other hand, allow for corrective actionbefore any harm is done. In the financial case mentioned, a leadingKPI was the number of planned customer meetings per comingweek. If the number of planned meetings fell below five per weekper employee, it would not be possible to maintain expected per-formance. With a six-week lead time this gave ample room forcorrective action.

In designing and redesigning production systems for produc-tivity and well being, the system designer must consider how thesystem should be monitored and managed. Management is notnormally included in the system design, but performance man-agement is. Since the performance management system influencesmanagement behaviour, it becomes imperative to discuss this aspart of the design process. It is not necessary for specific managerialbehaviour, but it is important to uncover the induced behaviour, orwhat can be called the unintended side effects of the performancemanagement system.

The effect of leading and lagging KPIs can also be translated intotwo management styles, as described in Blake and Mouton’smanagerial grid (1964):

1) Leading KPIs focus management on effort, and according to theconceptual frame of Blake andMouton, this is a person-centredapproach.

2) Lagging KPIs focus management on outcome, and according tothe conceptual frame of Blake and Mouton, this is a task-centred approach

In our study of a large financial corporation (Edwards andMøller, 2010), we found a mix of the two management stylesamong the best performing managers. This implies that the first-line manager combines attention to process results with personalaspects in order to manage difficulties in producing the outputneeded and to allow variations in individual performance overtime. This assures that such variations do not end with employeesblaming themselves for lack of competences.

8. Conclusion

The challenges of integrating ergonomic and human factor as-pects in company decision making have been addressed by manyauthors. In this paper, we present an approach that can be used byactors responsible for system design process. As demonstrated byreferences to research projects conducted in our department, this isa feasible approach, especially when a stakeholder orientation isdominant among the parties involved. The conceptual frameworkpresented here does not focus on specific tools and procedures. Itoutlines some of the basic aspects to be taken into considerationwhen involved in designing or redesigning a system. In this designprocess, it is important to distinguish between the role of designinga system design process, and the role of participating in a systemdesign team. In the latter, many of the issues addressed here aretaken for granted, as they present themselves during the process.But for the person responsible for designing the system designprocess, it is important to be aware of the limitations specified inadvance and the limitations established for the design team duringthe process. In this paper, we have (with reference to some of ourstudies) addressed basic issues of relevance for the personresponsible for the system design process. They are collected inTable 3. These limitations may decrease the possibilities for the

Table 3Basic issues to be addressed by persons responsible for designing the system (re)design process.

Basic conditions Issues to address Options for the personresponsible for the systemdesigner process

Boundaries and scope Boundaries defined forthe system to bedesigned/redesigned

Possibility of redefinitiondue to limitations of atoo narrow scope

Participants in designand redesign

Identifying stakeholdersand their roles

Identify all relevantstakeholders andclarify their roles

The character ofknowledge

Clarify the differentunderstandings(metal models) amongthe design group andcentral stakeholdersof the system to be(re)designed

Reach agreement onthe character of theknowledge of thesystem’s causal relationswithin the design team.Navigate context inorder to avoid beinglocked fast by theordered domain.

Performancemanagementand leadership

Search for the keyperformance indicatorsto be used in theoperations of the system

Try to clarify centralindicators for controlof system performanceso that the functioningof the system in operationand the challenges tobe met can be foreseen.

K. Edwards, P.L. Jensen / Applied Ergonomics 45 (2014) 26e3232

team to actually design or redesign a system with a significantlybetter performance in relation to productivity and well being forthe system actors. We have concentrated on the basic premises forthe design process. An elaboration of the approach is needed,including further development of tools to support the approach.

References

Alexander, D.C., 1994. The economics of ergonomics. In: Proceedings of the HumanFactors and Ergonomics Society Annual Meeting October 1994, vol. 38(10),pp. 696e700.

Beevis, D., 2003. Ergonomicsdcosts and benefits revisited. Applied Ergonomics vol.34 (5), 491e496.

Beyer, B., Holtzblatt, K., 1998. Contextual Design: Defining Customer-centeredSystems. Academic Press. SF.

Blake, R., Mouton, J., 1964. The Managerial Grid: The Key to Leadership Excellence.Gulf Publishing Co., Houston.

Broberg, O., 2010. Workspace design: a case study applying participatory designprinciples of healthy workplaces in an industrial setting. International Journalof Technology Management Vol. 51 (No. 1), 39.

Broberg, O., Edwards, K., 2012. User-driven innovation of an outpatient department.Work. ISSN: 1051-9815 41 (Supplement 1), 101e106.

Broberg, O., Hermund, I., 2004. The OHS consultant as a ‘political reflective navi-gator’ in technological changes processes. International Journal of IndustrialErgonomics 33 (4), 315e326.

Carayon, P., 2006. Human factors of complex sociotechnical systems. Applied Er-gonomics 37, 525e535.

Carayon, P., 2009. The balance theory and the work system model...Twenty YearsLater. International Journal of HumaneComputer Interaction 25 (5), 313e327.

Carayon, P., Smith, M.J., 2000. Work organization and ergonomics. Applied Ergo-nomics Vol. 31 (6), 649e662.

Dul, J., Neumann, W.P., 2005. Ergonomics Contributions to Company Strategies.Keynote address HAAMAHA.

Dul, J., Bruder, R., Buckle, P., Carayon, P., Falzon, P., Marras, W.S., Wilson, J.R., van derDoelen, B., 2012. A strategy for human factors/ergonomics: developing thediscipline and profession. Ergonomics.

Edwards, K., 2010. Performance management. in Danish. Personalechefen, 18e22.August.

Edwards, K., Møller, N., 2010. Psychosocial work environment and performance. In:Proceedings of the International Conference: Towards Better Work and Well-being. Finnish Institute of Occupational Health. Helsinki Feb. 2010.

Freeman, R.E., Wicks, A.C., Parmar, B., 2004. Stakeholder theory and ‘The CorporateObjective Revisited’. Organization Science Vol. 15 (3), 364e369,.

Henrick, H.W., 2008. Applying ergonomics to systems: some documented lessonslearned. Applied Ergonomics Vol. 39 (4), 418e426.

Jackson, M.C., 2000. Systems Approach to Management. Klüver Academic Pub-lishers, London.

Jensen, P.L., 2001. Human factors and ergonomics in the planning of production.International Journal of Industrial Ergonomics 29 (2002), 121e131.

Kensing, F., Blomberg, J., 1998. In: Participatory Design: Issues and Concerns.Computer Supported Cooperative Work, 7. (Klüver Academic Publishers),pp. 167e185.

Lewin, K., 2000. The field approach: culture and group life as quasi-stationaryprocesses. In: French, W.L., et al. (Eds.), Reading 7, pp. 112e113.

Mintzberg, J., 1998. Strategy safari: a Guided Tour Through the Wilds of StrategicManagement. The Free Press, N.Y.

Neely, A., Gregory, M., Platts, K., 2005. Performance measurement system design: aliterature review and research agenda. International Journal of Operations &Production Management 25 (12), 1228e1263.

Nielsen, A.P., Edwards, K., 2010. The paradox of lean in healthcare: stable processesin a reactive environment. Advances in Occupational, Social and OrganizationalErgonomics, 335e345.

Norton, D.P., Kaplan, R., 1996. The Balanced Scorecard: Translating Strategy intoAction. Harvard Business School Press, Boston.

Rasmussen, M.S., Justesen, T., Dohn, A., Larsen, J., 2012. The home care crewscheduling problem: preference-based visit clustering and temporal de-pendencies,. European Journal of Operational Research 219, 598e610.

Singleton, 1974. Man-Machine Systems. Penguin Hammondsworth.Snowden, D.J., Boone, M.E., 2007. A Leaders Framework for Decision Making. Har-

vard Business Review, pp. 76e79. November.Sundaram, A., Inkpen, A., 2004. The corporate objective revisited. Organization

Science Vol. 15 (No. 3), 350e363. MayeJune.Vink, P., Imada, A.S., Zink, K.J., 2008. Defining stakeholder involvement in partici-

patory design processes. Applied Ergonomics Vol. 39 (4), 519e526. July.Weick, K.E., Quinn, R.E., 1999. Organizational change and development. Annual

Review of Psychology 50 (1), 361.Westgaard, R.H., Winkel, J., 2011. Occupational musculoskeletal and mental health:

significance of rationalization and opportunities to create sustainable produc-tion systems e a systematic review. Applied Ergonomics 42, 261e296.