Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

Manufacturing operations are continually exposed to a large number of disturbances and fluctuations both from outside and inside the organisation that decrease their potential output. Since the ideal conditions for maintaining a stable, high-quality and low cost output in manufacturing are those associated with stability, manufacturing organisations strive for establishing such conditions by various techniques and ways of organizing.

One of the methods used in manufacturing history was standardising the product, a technique that became a real success in the beginning of the twentieth century. A good example of such thinking is the expression that is ascribed to Henry Ford concerning the Ford Model T: “Any customer can have a car painted any color that he wants as long as it is black.” Other examples of methods for creating stability in manufacturing are standardization of procedures, production levelling and Just-In-Time, all being essential parts of “lean production”. However, manufacturing systems are characterized by a variety of disturbances and fluctuations that must be managed. Under certain conditions these can be decreased by methods and techniques mentioned above. For a large majority of cases though, manufacturing firms use more or less advanced administrative systems to regulate and control their operations in order to handle the disturbances and fluctuations in an acceptable or favourable way.

Thus, control of operations is central in manufacturing and huge efforts are put on improving the efficiency and effectiveness of the control system. The complexity in the control task is high and the amount of data needed and used is extremely large. Together with human decision-making, Information Technology (IT) has been the tool for improving control for decades now. Software developments still lead to more and more sophisticated information systems, often termed Enterprise Resource Planning (ERP) systems. Newer ERP systems are modularized and incorporate a range of modules from finance and accounting to manufacturing and logistics, and some are even providing supply chain management solutions (Dery et al. 2006). These systems can cover the control needs of the whole organisation, from sales to manufacturing, delivery, and invoicing. The investments in ERP systems are high. The worldwide market for ERP systems has reached 79 billion US$ annually (Gefen and Ragowsky 2005).

These systems function very well in many applications and are believed to lead to significant cost savings and increased profitability by delivering operational advantages like reduced procurement costs, smaller inventories, more effective sales strategies, lower administration costs and reduced direct and indirect costs (Dery et al. 2006). Furthermore, they are expected to improve decision making because of their ability to provide specifically designed “real time” information to assist different management functions and procedures. However, they are also considered to bring unforeseen costs of large magnitudes as well as leading to unintended outcomes. Implementation costs often exceed cost savings and revenue gains. Between 60 and 90% of implementations failed to achieve the projected return on investment (Stedman 1999; Trunick 1999). These shortcomings during the 1990s have resulted in a growing literature on implementation success factors and better implementation knowledge. Wang et al. (2008) found that the consistency among the success factors had a significant positive impact. Ngai et al. (2008) identified 18 critical success factors and more than 80 sub factors for successful implementation of ERP systems in their literature review of 48 publications. They also concluded that it is not only implementation that determines the success. But the measurement of success remains very difficult for such projects. Some authors argue that the implementation should be seen as the first step towards a more or less continuous process of developing the fit between the operations and the ERP system. This post implementation process needs to be carefully designed to make the ERP system profitable (Nicolaou and Bhattacharya 2008). Yeh et al. (2007) concluded that the service quality of semiconductor-related industries in Taiwan was improved by ERP implementation. But they also remarked that the failure rate of the implementation of ERP systems in Taiwan is very high. In general, ERP implementations are still considered to be a high risk venture with uncertain outcome.

The design and functionality of ERP systems are such that they transform the nature, structure, and management of work throughout the entire organisation (Dery et al. 2006). End-users of ERP systems, being those who are dependent on their interaction with the software to fulfil their task, differ in needs and conditions for using the system. Calisir and Calisir conclude that there is a need to define user categories broadly in order to understand the end-user satisfaction as defined by the learnability and the perceived usefulness of the system (Calisir and Calisir 2004).

Dery et al. (2006) found that ERP publications are focussing mainly on the functional/technical aspects of the software and their likely impact on business performance in terms of efficiency gains, improved information flows, data processing, and profitability. Another main focus is laid on the implementation of ERP systems. Recently, some studies have been addressing the usage and maintenance of such systems. However, the literature is mainly managerially oriented, and there is a lack of studies that examine aspects related to work organisation and workplace performance as post-implementation responses, as well as effects on workplace control, power, and resistance (Dery et al. 2006). Moreover, these systems incorporate models of operations that were shaped not by the users, but by the designers of the systems – some of them reach back in history, as they are building up on older models during the development process – and these models may well be different from the work system where the ERP system is supposed to be used (Benders et al. 2006; Light and Wagner 2006; Locke and Lowe 2007).

Thus, the problem of getting value out of ERP implementations is not well understood due to shortcomings of current ERP research. From a professional practitioner’s point of view questions like “Why does the same IT system work differently in highly similar production systems?” and “Why does not a more sophisticated IT system automatically lead to more control?” are expressions of these shortcomings that need to be explained and further investigated.

In this chapter, we will suggest an alternative approach for elucidating the problems of the interaction between IT system and workplace performance. We present a view of the control system in manufacturing that is taking into account humans as actors for control. It is a view that considers control of a manufacturing system to be performed by a large number of actors in the system. It is furthermore a view that takes into consideration different means for control that are available. This implies not only using the technology for control, the IT system, but also using the organisational opportunities as well as the human capabilities of each employee.

The view is based on the socio-technical system approach and hence on the idea of considering a manufacturing system as a work system composed of a technical system and a social system, which interact and are mutually dependent. When changing the system (e.g. due to the implementation of an ERP system) the social as well as the technical system need be analysed and changed in accordance with each other and the overall goals of the change. Our position is thereby not to regard ERP systems as being unable to satisfactorily control operations, but to present how their potentials can be exploited by taking into account the power of human control. This means that the work system is considered to be controlled by the interplay of humans and technology. Control is executed in a distributed way in the system as a whole. Many different actors in the work system are – to some extent – involved in decision making with respect to control. A more elaborate discussion on how the chapter is related to socio-technical theory can be found in Sect. 10.2.6.

One characteristic of a work system that clearly demonstrates the difficulty of centralised planning is the fact that the degree of detailing is at its highest at short term shop floor level and decreases with distance in time from present (cf. Fig. 10.1). Since centralized planning is most likely not able to take into consideration all detailed information available at the short term shop floor level, control needs to be distributed throughout the organisation.

Fig. 10.1
figure 1

The temporal range of planning activities in production (modified from Scherer 1998)

Full size image

Control is thus constituted by human control behaviour of all humans involved in planning and scheduling activities - in one way or another, formally or informally. However, their control behaviour is conditioned by the availability of different kinds of resources, with the ERP system being one of the most important. In general, these resources are determined by the way work systems are (socio-technically) designed and organized. Any change (e.g. implementation of an ERP system) will consequently alter the available resources and hence the human control behaviour. If the change leads to a poor fit of the conditions and the demands for control, the control within the system will decrease instead of improve. This usually happens unexpectedly and control loops that were unknown up to date are revealed by changes.

An example of this is the machine operator whose schedule was updated daily instead of weekly after a change. His ability to locally optimise his machine concerning setups and loading was thus heavily impaired, because of frequent schedule changes. At next period’s degree of machine utilization assessment, he was blamed for decreased performance, although he had no longer any influence on the schedule and thus could not affect the performance. It was not until after the change that it was discovered that the implementation of the new IT hampers local control. This is because the implementation project focused on the IT’s ability to control rather then on the impact the implementation of the IT has on the humans’ ability to control – especially those humans who’s main task consists not of planning and scheduling are normally not considered at all.

Furthermore, research on control behaviour suggests that human control requires the notion or feeling of being in control. Consequently, the objective fit between the demands for control and the organisational resources is not sufficient to improve control. A subjective perception of being in control is also necessary to improve human control behaviour and hence system control performance.

The need for a personal feeling of being in control is recognized at the top management levels. Consequently most ERP systems provide sophisticated functions to instil the perception of being in control to management. A tempting strategy sometimes applied for this purpose is to (mis-)use integrated ERP systems for a very detailed data monitoring. The expectation is that in doing so, an in-depth description and analysis of production and logistic processes is generated and hence an improvement of operational control as well as perceived control is achieved. However, we assume that in many cases this strategy increases the management’s perceived control only, whereas its actual implication on the control capability of the work system remains unclear. There are several reasons for this assumption.

One is that managers suffer from the same human limitations as everybody else concerning the ability to use large quantities of data for decision purposes. Another is that the data in the system may not fully correspond to reality, which means that the detailed monitoring is done on a more or less invalid description of the real situation. Moreover, other effects of ERP system implementation may be that the flexibility of the work system decreases, that work design principles are violated, and that the basic ideas behind such an IT system and the way a company is organized may be in conflict (Benders et al. 2006; Koch and Buhl 2001).

A further problem is the increasing complexity of the ERP system itself. The objective of a complete IT-based model of the organisation makes the ERP system very complex and thus difficult to understand, to interact with, and to maintain. This development increases the demands on the humans and makes their work more difficult instead of facilitating it.

Against this background, we suggest a framework for assessing the control capability of a work system by outlining the conditions for human control behaviour and the resulting control determined by the socio-technical work system. We furthermore suggest how control capability can be analysed in order to predict the influence of different means taken to improve control, the most important one probably being implementation of new IT.

The chapter is arranged in the following way:

  • The second section outlines a system design model and its underlying perspectives.

  • In the third section, the behavioural perspective and its influence on the control capability of a work system is developed and underpinned.

2 Control Model

Control can be seen as a part of all coordination activities within a work system, with the goal to smoothen operations to enable order fulfilment as good as possible. From a technological perspective, the concept of control in work systems can be modelled using control theory for depicting the entire system (cf. Fig. 10.2).

Fig. 10.2
figure 2

The manufacturing work system depicted as a technical control system (modified from Scherer 1998)

Full size image

Figure 10.2 shows that the interplay between the information system and the physical system is highly integrated. The control input to the system consists of the objectives to be reached concerning output. Moreover, Fig. 10.2 gives information on sources of disturbances and unforeseen events. As can be seen these disturbances emanate both from within the system and from outside (Grote 2004), and they can be of informational or physical nature. Furthermore, the control system must also be able to handle the natural variances in operations as well as the variations imposed by the complexity of the product and production system as such (McKay and Wiers 2004). Disturbances, unforeseen events, variations, and variances together form the operational uncertainty of a work system. They all contribute to the control requirements, which can be defined as the sum of proactive and reactive actions required to reach the objectives. A more elaborate discussion of operational uncertainty and control requirements can be found in the next section of this chapter.

However, the model in Fig. 10.2 may be interpreted as a strictly technical model. This might lead to the wrong conclusion that control of manufacturing is a strictly technical issue, ignoring the fact that control of a work system necessitates people acting for control both within the information system and within the physical system. Consequently, control can be described as being composed by people exerting control in a network within the work system. In such a view, the control system is distributed and interlinked with the work system. However, the control system and the work system do not necessarily divide similarly into subsystems, i.e. one control subsystem may be interlinked with several work subsystems and vice versa.

Moreover, the control system as well as the work system is composed by mutually dependent technical subsystems and social subsystems. Control can thus be exerted by actors using the technical subsystem (mostly the ERP system), the social subsystem through direct interaction with other employees or – more likely – by a combination of both.

In contemporary manufacturing, the controlling actors are coupled by the ERP system, which often is supposed to be the overall coordination system. A problem related to this is that some couplings are tight and some are loose or very loose and in some cases the coupling does not exist. Furthermore, control actions are performed by human actors and each human actor has his/her own relation to the overarching coordination system (ERP system) and thus chooses to what extent his/her actions should be coordinated through the system. The coordination through the system may even violate the ability of the actor to fulfil his/her task. Research on scheduler’s work has been demonstrating that this problem is frequent, because apparently it is often a vital part of the scheduler’s activities to make sure that shop floor operators follow the schedule (McKay and Wiers 2004; Scherer 1998).

Another problem is that ERP systems sometimes cannot be configured to fit the reality. Interlinked bills of material or products that must be described with different units are often very difficult to handle. For example, a manufacturer using the same standard sheet metal for many different products may have difficulties in planning the need of the sheet metal since it is dependent on the mix of sales and not on single products. A sawmill might need to describe the same piece of lumber in the ERP system as a length (in meters), as a volume (in m3), as a part of a pile for drying which in turn is different from the same piece of lumber described as a part of a wrapped delivery pile. Another example is a low-value customized product in large volumes like a door or window that can be equipped with different customized fittings and delivered in any colour. If the ERP system requires a unique structure for each customized product, it will be very costly to control the customization through the ERP system. However, more recent ERP systems might offer a configuration module that can handle these problems.

In cases where the ERP system is not sufficient, schedulers often use self-made spreadsheets as an extension. Control exerted through the use of these spreadsheets is another example of control that is not or at least loosely coupled to the overall coordination system.

Control actions are thus performed more or less coordinated by the ERP system. How to describe the total control of the system is an unanswered question.

If the system is considered to be composed of a structure for control and of actors using this structure, the structure and the actors can be analysed separately concerning their contribution to the system control. Against this background, we suggest a process model for control basically consisting of six variables (cf. Fig. 10.3). In the following subsections these variables are described.

Fig. 10.3
figure 3

Process diagram describing the process of executing control

Full size image

2.1 Control Requirements

Control requirements are determined by the complexity of the subject matter of control. Ashby’s Law of requisite variability relates to actors trying to reduce the variety in output with the aim of keeping the system stable (Ashby 1957). According to Ashby’s law the actor’s variety is the only possibility to destroy variety in the system to be controlled. The law of requisite variety has two important consequences:

  1. 1.

    The amount of information available determines the amount of appropriate actions that can be performed.

  2. 2.

    To be able to control a situation the variety of the controller must be equal or greater than the variety in the system to be controlled.

The law of requisite variety thus describes that the variety of actions possible in the control system must as a minimum be as large as the variety of changes that need to be compensated in order to achieve control. The larger the variety of the controller’s options is, the larger the ability to cope with changes through regulation. A control system that is tightly optimised for a limited variety might be more efficient as long as the variety does not change. However, it cannot survive should the variety increase above the controller’s variety.

The law of requisite variety thus has important implications on flexibility issues because of its focus on possible outcomes within a set of defined situations. Consider for example the task of manufacturing various products with different assembly times at one assembly line. In this case, there is no possibility to adopt the manufacturing to the different demands of the products. Would there be at least two different assembly lines, the planner would have the possibility to control the manufacturing in a better way by splitting the products with long assembly times from products with short assembly times.

2.2 Control Opportunities

To be able to cope with operational uncertainties and to reach desired objectives, the actors need to be provided with control opportunities. Control opportunities are the sum of all available opportunities any actor may apply to perform any action of control. That is, control opportunities define the possible control actions available to an actor to change the state of the system. Control opportunities should permit the actor to reach the desired goals as well as to define means and standards to reach them. Control opportunities are determined by the structural properties of the system and can thus be deliberately designed to a certain extent. Information availability, organisational solutions, and technology are introduced below as being important parts of the structural properties.

Information availability concerns the type of information needed for the actor to perform the control task. In order to manage task related uncertainty and to reduce equivocality, organisations have to process information (Daft and Lengel 1986). The actor needs access to relevant, understandable, accurate, and timely information to be able to control the situation. Actors need relevant information such as information about the objectives, the status of the situation, the actions required, and the planned schedule.

Further, it is also necessary to look at different aspects of the organisation that may limit the ability of the actor to perform his control task successfully. The organisational solution will either enable or hinder the actor to act according to the demands of the situation. This has been a central tenet in the design of socio-technical solutions for planning and scheduling (Slomp and Ruël 2001). Examples would be the negotiation of additional manpower, a different shift pattern, or a resource changeover. As a result, it could be necessary to alter the authority of the actor and the needed support from the organisation. Within high uncertainty contexts it is more likely to observe control benefits when individuals or teams are able to make use of autonomy and collective problem solving than in low uncertainty environments (Parker and Wall 1998).

Finally, control opportunities are also depending on technology in different ways. There is the controlled system as well as the controlling system and its different control characteristics that must be considered. The complexity of these technologies and the extent to which they are understandable for the actors will affect many control opportunities.

To summarize, the control opportunities of a work system are the sum of all designed possibilities that the actors controlling the system can use in order to reach a system state that corresponds to the overall objectives of the system.

2.3 Control Skills

In order to make use of the control opportunities as described above, the individual actor must have adequate control skills. Control skills are thus defined as the ability of a human to produce control by acting within a certain situation using the given control opportunities of that particular situation. Professional knowledge of control methods as well as competencies to make use of control opportunities – such as available instruments and tools – are at the core of the actor’s possession of control skills.

However, detailed knowledge about the system to be controlled is also required. Hence, it is a fundamental precondition for skilful control that an operator has the necessary (tacit or explicit) knowledge of the cause-effect relationships that can be used for control. Based on the writings of Von Wright (1971), Petersen (2004) distinguished between what an actor is doing and what change she or he is bringing about. The actor performs control actions in order to bring about system changes. The desired change may however not be a direct consequence of the action but something it is brought about by a causal relation inherent in the controlled system (cf. Fig. 10.4).

Fig. 10.4
figure 4

The doing and bringing about aspects of control actions (Petersen 2004)

Full size image

The causal relationship between doing and bringing about might be easy to define and realise in physically proximal and simple control situations. However, the causal relation might also be mediated by the ERP system or other technical or organisational means, and there might be time delays involved that make the causal relation difficult to foresee or understand. Regardless of the relation, the control skill needed is depending on the content of the relation and the control agent’s knowledge about the relation and his/her ability to perform the actions needed.

Another fundamental aspect of skills in control tasks is what can be called situation awareness. This is a concept that was originally developed for aircraft pilots but that has spread to other domains. One of the most comprehensive definitions was provided by Endsley: “Situation awareness is the perception of the elements in the environment within a volume of time and space, the comprehension of their meaning, and the projection of their status in the near future” (Endsley 1988). For operators maintaining a system or operating complex equipment, situation awareness generally means the sensory, perceptual, and cognitive activity that prepares the user to make a decision, but it does not include the action as such (Pew and Mavor 2007).

The process of exerting control skills could thus be defined as – in the first step – assessing the situation to upgrade the situation awareness which would provide the control agent with the necessary cognitive background for making the right decision, and then – as a second step – continuing with doing what is necessary in order to bring about the desired change of the system. It may be noted that seldom all available control skills in a work system can be used. There might be control skills that do not correspond to any control opportunity.

2.4 Control Capability

As control capability we define the intersection of control opportunities and control skills, i.e. control capability is given when both, control opportunities as well as adequate control skills exist at the same time and the same place. This means that the total control capability of a work system can be regarded as the fit between the control opportunities and the control skills. Hence, control capability is the maximum control that can potentially be executed by the system (cf. Fig. 10.5). Whether or not this control capability results in control depends on two aspects. On the one hand needs to fit the control requirements. On the other hand it manifests in control behaviour only, when the humans are motivated. The following section elaborates on the relation between motivation and behaviour in control.

Fig. 10.5
figure 5

The relation between control opportunities (CO) given by all the structural possibilities for control of the work system (solid line), control skills (CS) given by the summarized abilities to exert control (dashed line), control capability (CC) which is the sum of all usable possibilities for control (shaded area), and the resulting control behaviour trajectory (CB), i.e. the utilized control

Full size image

2.5 Control Motivation and Control Behaviour

Control behaviour incorporates all control actions the human controllers actually perform. These actions are determined by the control opportunities and the control skills (i.e. by the control capability). Whether or not the human controllers make use of the whole scope of possible actions as determined by the control capability is dependent on their control motivation. Consequently we take into account that human planning and scheduling activities are determined by the objective means available for being in control (i.e. the control capability) on the one hand and the subjective control motivation on the other hand. As for example described by Scherer (1998), control decisions in practice are influenced by different rules. Some of them are defined by the control opportunities described above, but some are also determined by informal rules inherent in the social situation of the actor. These informal rules are depending on the cultural characteristics of the organisation.

The control motivation depends on emotional and motivational processes connected to the individual behaviour of each human actor. The decisions and actions taken can be seen as a behaviour trajectory, which describes the sequence of actions according to the actor’s actual motivation (cf. Fig. 10.5). The psychological processes and mechanisms that are underlying control motivation are further developed in Sect 10.3.

So far, the core of our process model of control has been outlined (cf. Fig. 10.3). It suggests on the one hand that control behaviour results from an interaction of control opportunities, control skills and control motivation. On the other hand it postulates that control behaviour needs to meet control requirements in order to maintain control. In the following, this model is elaborated further, mainly by differentiating the control requirements as well as the control behaviour, and by integrating feedback loops.

2.6 Operational Uncertainty

As described above, the law of requisite variety (Ashby 1957) assumes that the amount of variability in control behaviour should always exceed the amount of variety in the processes to be controlled in order to enable the system to cope with occurrences. These dynamic occurrences might be expected, unexpected or inexperienced and are causing uncertainties that have to be regulated. The quality and effectiveness of all activities concerning the regulation of uncertainties are critical for the outcomes and profits of a work system. In fact, already more than forty years ago uncertainties were seen as a fundamental problem for complex organisations (Thompson 1967). The ability to manage uncertainties successfully is one of the key factors to sustain and even expand a business. Uncertainty management therefore has become a research interest. Consequently, the amount of uncertainty embodied in the control requirements defines the variability in control behaviour that is necessary to control the situation.

In planning and scheduling, a huge variety of uncertainties needs to be regulated. These uncertainties can be due to many different contributing factors. Generally it can be differentiated between external and internal factors. External factors are situated outside an organisation. Typically, market dynamics fall into this category. Examples are the limited predictability of customer demand as well as unexpected changes in existing orders. However, there are many more uncertainties caused by external factors: unreliability of suppliers, supplies with bad quality, new laws and other legal prescriptions, strikes and other occurrences suitable to disturb logistic processes, just to name a few.

Uncertainties can also arise internally. Examples of such uncertainties are technology and production process related uncertainties, distributed information, information loss at organisational interfaces, missing or ambiguous specifications and instructions, unavailable key persons, lack of rules or standards for collaboration, machine breakdowns, and bad product quality which causes additional effort.

Furthermore, as planning decisions are highly interrelated, even planning decisions themselves can potentially lead to uncertain situations. Today’s solutions may easily implicate tomorrow’s problems. Sometimes these problems do not affect the original decision-maker, but somebody subsequent in the stream of actions. This does not necessarily require improvidence or even bad intentions of the original decision-maker. It might happen just because of ignorance towards system complexity. Maybe the original decision-maker did not at all have a chance to appraise the impact of his decisions on other people’s work. Hence, the mere design of planning and scheduling structures and processes including the allocation of planning and scheduling tasks can be a source of uncertainties (Grote 2004).

In particular, there is a dilemma regarding allocation of planning and scheduling tasks to centralized planning departments and the shop floor respectively (Wäfler 2001). Allocating autonomy to the shop floor causes uncertainties for the planning department, which only with immense effort can keep an up-dated picture of the shop floor’s dynamics and therefore runs into the danger of planning on an outdated information basis. On the other hand, allocating as much decision-making authority as possible to the centralized planning department is not well suited for taking into account detailed local information when making planning decisions. Consequently a well-balanced allocation of decisions-making competencies among planners and schedulers as well as between the shop floor and the planning department is required. In order to allow for flexible (re-)acting on uncertainties, it might also be required to design the allocation of decision-making authority in an adaptable way (cf. below ‘structural control’).

Figure 10.6 summarises the consequences of the statements made so far for the control model: in planning and scheduling, control requirements (CR) incorporate operational uncertainties (OU), which arise from environmental dynamics (ED) as well as from the internal structures (IS) of planning and transformation processes.

Fig. 10.6
figure 6

Internal structures (IS) and environmental dynamics (ED) are defining operational uncertainties (OU) which lead to control requirements (CR)

Full size image

Since the control model depicted in Fig. 10.6 suggests that the scope of required control behaviour should fit the control requirements given in a concrete situation, a measure for control requirements or for operational uncertainties would be very helpful when designing planning and scheduling structures. However, well defined and operationalized propositions for conceptualisations of operational uncertainties are rather rare in literature.

Control of operational uncertainties is a core issue in the socio-technical system design approach (for a comprehensive description of this approach cf. Van Eijnatten 1993). This approach takes into account that business organisations consist of a technical as well as a social sub-system. The former incorporates all machines, technical resources, prescriptions, regulations, condition set by the factory layout, and the like. In contrast, the latter incorporates the humans as individuals as well as groups of humans with their needs and behavioural patterns. Since the two sub-systems are following different logics, the socio-technical system design approach postulates that only a combination of the two allowing for a joint optimization is suitable for the successful functioning of the system as a whole. Proponents of the approach criticize that in practice mostly a one-sided optimization of the technical sub-system can be observed. This is especially true with regard to the design of planning and scheduling processes, which mostly are driven by IT-implementation. However, such a one-sided optimization is considered to be sub-optimal not only since it runs the risk to hamper human flexibility and creativity required for a successful mastering of operational uncertainty. Moreover, it might even destroy respective human competencies because it deprives employees of working conditions allowing for the development of required know-how, experiences, and motivation.

However, the socio-technical system design approach considers the successful functioning of a system as a result of competent coping with continually changing conditions and hence with variances or disturbances with respect to the plan.

In general, the socio-technical system design approach favours a system design that allows for a local regulation of variances and disturbances. Therefore, it promotes a decentralization of decision-making into autonomous organisational units, requiring as little cooperation as possible at the organisational interfaces between these units. However, applying this concept to the design of planning and scheduling structures proofs to be rather difficult (Wäfler 2001), since it presupposes that the task can be modularized and allocated to organisational units in a way making the units independent of each other. In contrast, planning and scheduling incorporates coordination of the order flow through the organisation and hence requires regulation at the interfaces between organisational units. Therefore, with regard to planning and scheduling, organisational units remain dependent of each other. Often it is aimed at mitigating this dependency by introducing buffers at the interfaces. However, such buffers often increase lead times and therefore hamper efficiency.

In fact, the socio-technical system design approach gives hints on how to deal with uncertainty (by allowing for local regulations on the basis of autonomy and independency). But it does not really define what uncertainty is or how it could be measured. It considers every variance or disturbance of the normal or planned workflow that could critically influence the performance of an organisation as an operational uncertainty. However, this definition of uncertainties is very general and does not provide detailed information about the features or the measurement of occurring problems in the planning and scheduling context.

Wall et al. (2002) propose a definition with implications for measurement. They suggest that operational uncertainty is defined in terms of the number and difficulty of problems, key variances or exceptions that have to be accommodated. This definition refers to problems as countable entities of a specific source with a certain “difficulty” for their regulation.

In addition, some definitions highlight further psychological aspects of uncertainties and their complexity. Jackson (1989) for example defines operational uncertainty as a lack of knowledge about production requirements like the occurrence of problems and how best to deal with them. Wall et al. (2002) illustrate that this concept represents a lack of understanding about cause and effect between knowledge and uncertainty. Where such uncertainty is high, knowledge is incomplete and therefore problem-solving requirements are high.

Milliken (1987) distinguishes between three dimensions of uncertainty based on lacking information: state uncertainty (concerning the future development of the environment), effect uncertainty (regarding possible effects of changes on the organisation) and response uncertainty (regarding alternative responses and the prediction of their consequences).

With respect to the work task, Clegg et al. (1989) describe task uncertainty as a lack of predictability over matters as the timing of, and demand for, a particular task, the meaning of inputs that trigger the need for a response, and the nature of the required actions or responses.

Summing up these views, operational uncertainties can be considered as situational changes caused by occurring variances and disturbances. Uncertainties are characterized by a lack of knowledge about the new situation. The lacking information can refer to the “when” and “why” of an uncertainty (constrained predictability) and/or to the further development of the situation and the effects of possible actions (constrained transparency). Uncertain situations are triggered by external or internal variances and disturbances and need to be regulated adequately in order to avoid or decrease negative outcomes. For that purpose control opportunities and control skills are needed which allow for control behaviour. The following section describes structural and operational control as different forms of control behaviour.

2.7 Structural and Operational Control

A crucial premise for successful coping with an operational uncertainty is the fit between control requirements (as consequences of operational uncertainties) and control behaviour. This control behaviour is a result of control opportunities, control skills, and control motivation (cf. Figs. 10.3 and 10.6). However, two different kinds of control behaviour can be identified: structural and operational control (cf. Fig. 10.7). Whereas the latter refers to control behaviour within given structures, the former refers to changing the structures themselves. Hence, the structures of an organisation as defined for example by organisational and job design, by provided instruments and tools, and by the human planners’ qualification and skills provides the conditions for the operational control behaviour that can be deployed to cope with uncertainties. Structural control surmounts such limitations. It refers to changing given structures and therefore for changing limitations of operational control.Footnote 1 In the following, these two types of control behaviour are described in more detail.

Fig. 10.7
figure 7

Control behaviour (CB in Fig. 10.6) is differentiated into two types: structural control behaviour (S-CB) and operational control behaviour (O-CB). The latter refers to all actions aiming at coping with control requirements and hence is directly influencing ‘resulting control’. However, O-CB is determined by an organization’s structural conditions. The structure limits the scope of possible O-CB. Changing such structural limitations requires a change of an organization’s structure. S-CB refers to such behaviour, resulting in a change of structures and hence modifying the frame of O-CB

Full size image

Control opportunities are possibilities to influence a situation in order to reach certain goals, to set these goals as well as to define the means and regulations to reach them. Control opportunities can be designed to a certain extent. Grote (2004) proposes to design the control opportunities in a way balancing dependency and autonomy (cf. Fig. 10.8). Dependency results from a centralized planning system that reduces local degrees of freedom. Such design is aiming at avoiding uncertainties by feed-forward control. The overall aim of plans is to minimize uncertainty by preventing from or (at least) by preparing for future uncertainties. This is achieved for example by an exact timing of production processes or by reservation of production resources. In such systems, plans are prescriptions to be followed. Deviation from the plans would cause additional uncertainty for the centralized planner. The prescriptive character of the plans creates dependency. The problem of such a strategy is its inflexibility to react to unforeseen events. However, the more complex production becomes, and the more dynamic markets are, the more likely is the occurrence of unforeseen events.

Fig. 10.8
figure 8

Basic principles of uncertainty management underlying organisation design (adapted from Grote 2004)

Full size image

In order to be able to cope with such unforeseen events, feedback control is necessary. In line with the assumptions of the socio-technical system design approach (see above) effective feedback control requires opportunities for local regulation of uncertainties. Consequently, local independency and autonomy is needed in order to allow for local coping with unforeseen events and hence for preventing uncontrolled spread of problems throughout an organisation. In such systems, plans are not prescriptions but rather resources for local acting by providing transparency and orientation.

As mentioned above, Grote (2004) suggests an organisational design – and hence a design of control opportunities – that balances autonomy and dependency by a loose coupling of organisational units. However, she also pleads for flexibility regarding the balance of autonomy and dependency. Such flexibility should allow an organisation to flexibly adjust its organisational mode (i.e. its design of autonomy and dependency) in order to allow for an optimal coping with actual variances and disturbances.

Control of the organisational mode is a core aspect of structural control since it involves the (re-)assignment of decision-making responsibility. Dependent on the organisational mode chosen, more or less decision-making responsibility is assigned to the shop floor or to the centralized planning department. However, structural control involves more than controlling the organisational mode. It involves all behaviour that influences an organisation’s structure and hence also its control opportunities and control skills (cf. Fig. 10.9). This includes activities like designing products and production, designing the control system and especially the control tools and instruments, but also qualifying the humans in order to develop their skills.

Fig. 10.9
figure 9

Structural control changes internal structures (IS) in order to optimize control opportunities (CO) and control skills (CS). However, the internal structure also influences operational uncertainties (OU)

Full size image

Figure 10.9 also shows another important property of structural control behaviour. It does not only influence an organisation’s overall control ability by determining control opportunities and control skills. It also influences the control requirements. This is due to the fact that operational uncertainties are partly depending on an organisation’s structure, which in turn is influenced by structural control behaviour. De Sitter et al. (1997) argue that a suitable structuring of production process can reduce control requirements up to 80%. In their concept, “suitable” mainly means parallelization and segmentation of production in order to avoid dependencies at the organisational interfaces. Thereby organisational interfaces are to be designed in a way reducing the need for boundary spanning cooperation as much as possible. Although total independency is not possible (see above, Grote 2004), structural design creates more or less dependency and hence more or less operational uncertainty and control requirements. With reference to planning and scheduling, the following are examples for such homemade control requirements (a) two planners need to access the same production resource, (b) continuation of production is delayed due to a missing approval from external persons (e.g. quality managers or constructing engineers), (c) centralized planners lack detailed situational knowledge and decentralized dispatchers lack overview, (d) local optimization is no more possible since the global optimization works in real-time and therefore creates uncontrollable local dynamics.

So far, a process model for work system control has been outlined (cf. Fig. 10.9). It defines a works system’s control capability as an overlap of control opportunities and control skills (cf. Fig. 10.5). However, whether or not the actors within the system are carrying out the potential control behaviours determined by the control capability is depending on their motivation. We understand actual control behaviour of humans as trajectories within the control capability field of possible actions. The following section describes psychological processes that influence these trajectories.

3 Emotional and Motivational Influences on Control Behaviour

In this section, we will explain important emotional and motivational processes, which are underlying individual behaviour. We will show how these mechanisms are connected to human control behaviour and hence to resulting control (cf. Fig. 10.9). The last part consists of proposals for improvement of technical innovation projects and for measurements of work design, which are concerned with the ability of the members of a work system to cope with critical situations.

3.1 The Role of Individual Control and Competence in Human Behaviour

3.1.1 Individual Control and Competence

When humans learn how to meditate, at first they have to learn to stop thinking. It is very difficult to stop the inner monologue. It is also very difficult or even impossible to stop emotions. Humans are always in a certain emotional state: angry, happy, hateful, calm… This emotional condition influences our whole stream of activities. Being angry means to be concentrated on the object of anger, it means also to be full of tension and ready to act. Possible actions taken into account have aggressive character. Happy people don’t have this readiness to act, they think about positive things and they are rather absentminded than concentrated. Happiness and anger are embedding our thinking and acting in a different way, leading it in a certain direction and influencing the quality of the results. By the way, the probability that extreme anger and extreme happiness are leading to good performance is low. Accordingly, the control behaviour trajectory described in our model (cf. Fig. 10.5) is strongly influenced by the emotional state of the humans.

Along to most psychological theories, the most important variable triggering our emotional state is individual control (Lazarus and Launier 1978). Individual control from a psychological point of view is the actual probability of a human to influence a situation according to his or her own goalsFootnote 2. If individual control has the value “1”, one has the perfect ability to reach all intended goals. If individual control is “0”, there is no chance to reach the goals at a certain moment. Different values of individual control are related to certain moods and behaviour patterns. Individual control close to zero for instance leads to a behaviour pattern which is called “learned helplessness” (Seligman 1975). If this is the case for a longer period, human beings and highly developed animals loose the ability to act autonomously and begin to stay in a state of complete inactivity.

The “classical” reaction to an actual loss of individual control is an increase of arousal, which is a biologically useful reaction. The organism becomes more activated and thus ready for fighting or fleeing, depending on what is most useful in the actual situation. Humans, especially in the context of modern work environments, typically do not react in that way. The normal reaction in such situations is unhealthy stress, because the arousal cannot be transformed into fighting or fleeing.

In situations with loss of individual control people can react in different ways, also at work, where individual control is to a large amount determined by the design of the human’s work task. However, the actual emotional state is not only influenced by the amount of individual control determined by the design of the work task. It is also influenced by the importance the work task has in relation to the individual’s goals, and by other aspects not necessarily related to work. There is a different effect on the emotional state when a very important part of work has gone bad in a situation with additional private problems than in a situation of perfect happiness. Therefore it is necessary to distinguish between ‘individual control’ as related to the actual work task, and a much more general belief in one’s own ‘actual competence’ (Dörner 1999), related to the estimated success to cope with all actual and future demands. According to most theorists, humans have a general motive to acquire such competence. This motive causes curiosity and empowers for learning.

Actually, subjective beliefs regarding ‘individual control’ and ‘actual competence’ are rather influencing a human’s behaviour and his/her emotional reactions than an objective evaluation of the two. Furthermore, individual control and actual competence are connected, especially if work is an important part of one’s life. However, the same disturbance in a work system can provoke very different individual emotional reactions, because the actual competence of its members can be different. Each disturbance that leads to a decrease of individual control also leads – differently for different individuals – to a loss of actual competence.

3.1.2 Control and Competence in Complex Work Situations

Psychological research describes typical patterns of behaviour which are related to impaired performance in very complex situations (Dörner 1996; Jansson 1994). Most of them are dysfunctions in the use of information during the process of planning and acting, for example:

  • ‘Encapsulation’ is the tendency to avoid problematic fields of action and to concentrate on aspects where great skills and competencies exist, disregarding the importance and necessity of these activities.

  • There is also a tendency to use mental models that are too simple for the adequate description of actual problems (e.g. only one reason for all problems). Hence, one’s cognitive limitations might be underestimated.

  • A typical mistake in prognosis is the tendency to believe that major actual constraints of a complex situation will not change in future.

Such critical behaviours happen in very different professional fields (engineering design, management, spatial planning), and they also happen to people knowing about the risks to act in that way. Dörner (1996) calls these and other similar behaviour patterns ‘intellectual emergency reactions’. They are typical for complex work tasks where the relation between human acting and its outcome is not clear. In such situations, the outcome of acting can be interpreted in different ways (e.g.: management strategy and success; product design and reaction of the market; the real quality of some typical regional planning activities can be assessed twenty years later).

In situations with non-existing, unclear, or long term cause-effect relations, it is possible to select specific information which allows the human to perceive a high amount of individual control although real individual control is low (control illusion). Encapsulation allows to avoid negative information by concentrating on own strengths, and hence leads to overestimating one’s individual control. Too simple mental models strengthen the belief in the own ability to understand the complexity of a problem, and so on. Research results and case studies show that a low level of perceived actual competence increases the probability of these intellectual emergency reactions because they satisfy the motive to feel competent for the actual demands. Under certain conditions the intellectual emergency reactions lead to a feedback loop which is amplifying itself concerning its negative consequences (cf. Fig. 10.10).

Fig. 10.10
figure 10

Positive feed back loop with negative consequences (please note: ‘actual competence’ is always the subjective perception of one’s own competence)

Full size image

According to this model, a low level of perceived actual competence leads to intellectual emergency reactions. The consequence of this behaviour is an inadequate mental model leaving out important information to cope adequately with the complex work task. This lowers performance, and coming from that decreases actual competence again. This positive feedback loop can have very important negative consequences.

In a work system, disturbances may decrease individual control (see above). However, also changes and innovations may decrease individual control. Hence changes and innovations may increase the risk of inadequate problem solving behaviour. This is especially true for people with a belief of low actual competence, because they are more likely to show intellectual emergency reactions.

As a consequence, the vicious circle depicted in Fig. 10.10 can decrease the resulting control of a work system (cf. Fig. 10.9), since inadequate problem solving behaviour is most likely to lead to inadequate control behaviour.

3.2 Competence Regulation in the Context of Teams and Organisations

The model of competence regulation was originally related to individual acting. In industrial contexts normally people have to work within groups and organisations and this leads to division of labour. Work is related to and depending on others to achieve a final collective result. Moreover, its consequences are not under individual control of the acting person. Therefore individual control depends on the control capability of the whole work system. Competence and control are generated also by additional factors, e.g.

  • Confidence in the regularity of work processes

  • Confidence in the organisation’s ability to cope with extraordinary demands

  • Trust in the personal qualities of the management

  • Trust in cooperation partners (colleagues, customers, suppliers)

Confidence and trust normally are generated by former experiences, i.e. typical behaviour in similar situations and resulting knowledge about resulting processes. But if people have to cope with complex and new demands, there is – by definition – not much knowledge about former processes and organisational resources. So people have to estimate the control capability of their teams and organisations in another way. This estimation in many cases is based upon trust in the general abilities and integrity of cooperation partners and leaders. This is the reason for the common experience that coping with major changes and complex disturbances in work systems depends on transparency and an atmosphere of trust. This is described broadly in management, leadership and organisational development literature (e.g. for ERP system implementation by Kwasi 2007; Holland and Light 1999; Krause and Gebert 2005; and on a more general level in this book by Günter et al. in Chap. 5). If trust and confidence is low, the positive feedback loop which has been described for individuals in Fig. 10.10 can also lead to a crisis of the whole work system (cf. Fig. 10.11).

Fig. 10.11
figure 11

Individual emotional processes and control capability. LP Loss of performance, AC actual competence, IER emotional emergency reactions, IMM inadequate mental models

Full size image

Loss of confidence lowers the level of perceived individual control and actual competence. This increases the probability of intellectual emergency reactions. In turn, that leads to more mistakes, lower performance, and less trust and confidence. The consequences are positive feedback loops leading to very negative developments and results.

These considerations may show the impact of emotional regulation processes on the control capability and performance of the whole work system. To understand fully how work systems cope with complex demands, emotional and motivational processes have to be integrated into the analysis. We will show this by the following examples.

3.2.1 Examples

3.2.1.1 Example 1: The Role of Fear in ‘Technical’ Innovation Processes

Implementing new software often leads to major disturbances and new structures. Work and business processes, work tasks, and the relations in teams are changing. Sometimes technical change can lead to loss of job. All this causes increasing uncertainty, and coming from that a loss of individual control. In many cases, implementation strategies do not consider these emotional aspects (von der Weth and Spengler 2007). The resulting behaviour patterns are, amongst others:

  • A negative attitude to the change process, because existing qualification is partly not needed anymore

  • No discussion about problems and mistakes, because people fear to be regarded as incompetent

  • Risk avoidance in learning to handle the new software, because experimenting with the new software possibly generates loss of face by mistakes

  • Strategies for individual improvement of the own work process, not regarding the demands of cooperation within the work system

  • Intellectual emergency reactions because of decreasing control

Behaviour of this type decreases the actual performance within the implementation process and influences the individual control as well as the emotional state of the implementation project team. In turn this leads to a higher probability of wrong planning and decision making of these people and diminishes the probability of success.

3.2.1.2 Example 2: Conflict Escalation

If control capability of the work system and individual competence is perceived as low, the quality of conflict resolution in the work system decreases. This can be exemplified with a common conflict between a new superior and members of his/her team. A new superior often feels the challenge to establish new processes and forms of communication because he/she wants to leave his/her marks on the team. The result of these activities is highly correlated to his/her own perceived competence in the new job. Any objections in this situation often lead to stress and a decrease of perceived control. This can cause aggressive or fearful reactions.

On the other hand we also have a critical situation for the team members. Their perceived competence is low because the innovative ideas of the new superior cause uncertainty vice versa. This leads to a positive feedback loop with decreasing control on both sides and can cause aggressive behaviour. Typically, both parties increase their efforts and involve allies in the organisation if they want to enforce their goals. In situations of that type, conflicts can start from minimal disturbances in the work process and spread like wildfire, causing serious damage in an organisational culture.

Escalations like this can happen in every work system, but their probability increases when the basic level of the work system’s control capability is low. A low control capability does not contribute substantially to the work system members’ perceived competence. On the other hand, a high control capability can be a source of trust and confidence and can support the individual competence of the parties involved in a conflict. With a higher perceived competence on both sides in the beginning of the conflict, the chance increases to solve the team conflict before an aggressive and irreversible escalation starts.

3.2.1.3 Example 3: Trust, Communication, and Knowledge Development

Long term developments in work systems are also related to emotional processes. The control capability of an organisation depends strongly on its knowledge. Motivation and strategies for individual knowledge acquisition are connected to the emotional state of the work system members. Like knowledge management on the level of organisations, individual knowledge handling strategies can be described in the terms of the knowledge management strategy of Probst et al. (2003): knowledge identification, knowledge acquisition, knowledge development, knowledge sharing, use of knowledge, and knowledge preserving. The quality of these activities is also connected to the control capability of the whole organisation. If the perceived individual competence is high and connected to an actually high control capability of the whole organisation, then the readiness to share knowledge in many cases is also high: The members of the organisation have seen many times before that sharing knowledge and collaborative work are useful. In addition, the members of the organisation have the chance to learn appropriate strategies for cooperative knowledge management for the benefit of the whole work system. This increases the control capability, because more integrated knowledge structures exist. Work systems of this kind have actual and intensively used information infrastructure.

Low perceived control capability of the work system is connected with many forms of fear, e.g. loss of job and payment. A more aggressive organisational climate is probable. In that case we have two phenomena which are counterproductive for knowledge development in the work system:

  • Fear prevents an open discussion of problems and mistakes, which is necessary for a good quality process and effective innovation. If people don’t share knowledge about mistakes and problem solving processes, they cannot learn from each other and the quality of knowledge and hence control capability decreases.

  • Fear also causes a behaviour which is called “knowledge retention” (Hacker 2005): If people are in a very competitive situation with their colleagues sharing knowledge is not useful individually, because it improves the chances of the competitors. Individual knowledge retention also has an influence on knowledge infrastructure. Tools are not maintained and hence loose actuality. Because of that, they are not used anymore and become more and more irrelevant. This also leads to decreasing control capability of the whole work system.

4 Conclusions

This chapter reflects on the complexity of control capability. It addresses the subject matter from two points of view:

  • A process oriented design model of control (cf. Sect. 10.2)

  • Emotional and motivational influences on control behaviour (cf. Sect. 10.3). The former refers mainly to objective aspects of control capability. At its core it defines a work system’s control capability as an overlap of control opportunities and control skills. This overlap limits the scope of possible control behaviour, and hence of resulting control. Resulting control is high, when it fits with control requirements as determined by operational uncertainty. However, control behaviour is differentiated into two types: operational control behaviour and structural control behaviour. Operational control behaviour on the one hand aims at immediate coping with operational uncertainties and hence has a direct impact on resulting control. On the other hand structural control behaviour influences an organization’s structure and hence the frame, which is determining the limits of operational control behaviour. Consequently structural control behaviour does not have an immediate impact on resulting control, but changes the frame of operational control behaviour and therefore has an impact on resulting control mediated by operational control behaviour. Such structural control behaviour allows for feedback loops in our process oriented design model of control.

However, as all human behaviour also control behaviour is influenced not only by objective aspects but also by subjective motivational and emotional processes. Control capability solely determines the objective frame for operational and structural control behaviour. No behaviour out of this frame is possible. However, not all the behaviour possible within the frame must necessarily become manifest. Subjective motivational and emotional processes influence the manifestation of concrete control behaviour.

Both, the objective given frame of control behaviour as well as the subjective processes influencing the manifestation of actual control behaviour need to be considered for understanding control. Following the implications of both for practice are discussed.

The main implications of the design model of control (cf. Fig. 10.9) are the following:

  • Control results from an interaction of control opportunities, control skills, and control motivation on the one hand and control requirements on the other hand.

  • Control behaviour can affect operations or structures. The former results in process control, the latter results in structural changes and hence influences control opportunities as well as control skills. Control therefore is self referential.

  • Furthermore, since control requirements are not independent from structures, structural control even impacts operational uncertainties.

  • Any change (as e.g. the implementation of a new ERP-system) is a structural change and therefore impacts control opportunities, control skills as well as operational uncertainties directly. If these changes result in an inadequate fit of control behaviour and control requirements, the resulting control is low although the new IT is better than the old one. An inadequate fit can be caused for example by the following consequences:

    • New regulation requirements might occur (e.g. due to change in information or control structure like required fixing of dates)

    • (Structural) control opportunities might be lost (e.g. due to proprietary software)

    • (Structural) control skills might be missing (e.g. due to lack of experience of in-house IT staff)

    • (Operational) control opportunities might be lost (due to centralisation, loss of transparency like daily or intra-daily changing job lists)

    • (Operational) control skills might be lost (e.g. because an experienced planner’s know-how is not applicable any more)

The main implications of our reflections on emotional and motivational influences on control behaviour are the following:

  • Individual thinking and behaviour are influenced by emotional processes.

  • The emotional state is influenced by the belief to be competent (to have actual competence). This belief includes individual control of the actual situation and more general competence for future tasks and demands. High belief in ones own competence is connected to positive emotions, low actual competence to negative emotions and stress.

  • A low level of competence is closely related to certain behaviour tendencies. The classical biological reactions are aggression or flight. Modern reactions especially in complex and new situations are so called intellectual emergency reactions (see above). In many cases, the resulting behaviour has the consequence that only positive information is selected and actual competence can be perceived as high, although it is low for the objective demands. Fear, aggression, and intellectual emergency reactions on the one hand and positive emotions on the other hand influence the behaviour trajectory substantially.

  • This has also an influence on the control capacity of a whole work system, its work processes and results. Negative emotions generate inadequate behaviour on the individual level and decreasing control capability on the work system level. There is also an influence of the control capability on emotional and motivational processes. In normal work situations, individual performance depends to a certain degree on the performance of the whole work system. Therefore, the control capability of a work system is an important source for the individual competence to cope with work related problems (e.g. innovation, disturbances).

  • A low control capability increases the probability of negative emotions and intellectual emergency reactions. This interdependency between individual competence and control capability can be the starting point for dangerous developments in an organisation, described in several examples of positive feedback loops in this chapter. Especially in situations with great uncertainties on the individual as well as on the work system level, the danger of such aberrations is high because uncertain demands – like technical and organisational innovation or disturbances – lower the actual competence of the members of a work system. In situations like that, it is necessary to act psychologically in a careful way. Project managers and superiors should avoid behaviour, which causes additional loss of control.

Our reflections have the following implications for the design of planning, scheduling, and control systems:

From a process oriented design model perspective as well as from a psychological perspective, planning, scheduling and control is complex, incorporating multiple feedback loops. Therefore the complex interplay of humans, organisational structures, and technology needs to be carefully considered when designing or changing the control system. It is especially required that the humans – who are the actual ‘producers’ of control – are empowered to really perform control. In general, this presupposes the following:

  • Objective conditions empowering the humans need to be provided, especially a control capability (as a result of control skills and control opportunities) that fit the control requirements. This allows for required structural and operational control behaviour.

  • To assure such objective conditions, comprehensive instruments for work system analysis and design need to be applied. These instruments are still to be developed.

However, as discussed in the section on motivational and emotional influences on human behaviour, the provision of the objective conditions is far from being sufficient. What is required in a complementary manner to the objective conditions is a belief of the humans to actually be in control. The following attributes of work systems support individual development of such beliefs:

  • Clear processes and tasks which are adapted to the individual’s performance level

  • Tasks which stimulate learning processes and interest in interdependencies within the work system

  • Information about processes in which the individual’s work is embedded

  • Chance for participation in activities improving the control capability of the whole work system (Kaizen, participative knowledge management)

  • An organisational climate reducing fear by productive procedures and methods for learning from errors and mistakes

Moreover, when changes occur (e.g. due to the implementation of new IT), the beliefs of being in control are endangered. For change management this means for instance:

  • Clear defined goals, quality criteria, and milestones for change projects

  • Participation in work design as early as possible in the process

  • The members of the work system should be informed clearly about future tasks and demands

  • The innovation process should be connected with qualification programmes which are adaptable to the specific individual qualification level

  • Emotional development should be observed carefully on individual and work system level. There should be “sensors” in the work system for fear, aggression, and intellectual emergency reactions