19th International Conference on Production Research
RECONFIGURABLE ORGANISATION TO COPE WITH UNPREDICTABLE GOALS Mario Sergio Salerno Production Engineering Department, Polytechnic School, University of São Paulo Av. Prof. Almeida Prado, travessa 2, n.125
05508-900 São Paulo – SP, Brazil
[email protected] Abstract The paper proposes a methodology and design rules for organizational structures facing higher necessity of internal adaptation, of rapidly reconfigure themselves to cope with unpredictable situations – new markets, new products, unpredictable mix of production, problems in production process or flows etc. It implies changing and often conflictive criteria for production goals and for the allocation of work. The methodology was developed based on a large field action research and consulting during late 90s’; it was discussed and applied in organisations like Unilever (food and consumer goods), Rhône-Poulanc (chemical), Rockwell (autoparts / mechanical), Petrobras (oil refinery & engineering), R&D sectors of some companies as well as some attempt in public sector, and others. Their basis is the design of auto-reconfigurable working groups – or groups with variable geometry, depending on the events to face. Groups have neither previous size (number of components) nor predefined composition. Keywords: Reconfigurable organisation, open groups, organisational design, flexibility, efficiency.
1 INTRODUCTION Contemporary operations efficiency must be closed linked to companies’ strategy. Many businesses are run in unpredictable and turbulent environments – new competitors, new technologies, new products, new financial conditions etc. Frequently it means that a production system must be designed and implemented without a clear vision of the range or mix of products it will be asked to deliver. Decision rules for production scheduling can change according to the situation – not only cost, sometimes product performance, sometimes time to market, sometimes a combination of them. Automation, low inventories and outsourcing complicate the scene: integrated production systems are more suitable to export varieties or problems to other workplaces, sections, departments, plants or companies; automated processes, with remote control and guidance by direct workers, require an abstract ability to be run, moreover an ability to face to unpredictable situations. Clearly, production structure must be designed to cope with the contemporary need of flexibility. Organisational literature is rich in discussing flexible schemes compared to the traditional tayloristic / fordistic ones. The main approach in that sense, the so-called socio-technical systems [1], [2], is a real advance but it has some conceptual problems and mainly it lacks a tooling kit to orient organisational design – the design rules. The paper proposes a methodology and design rules for organisational design to companies that faces competitive and technological environments like the one discussed before. To do so, a brief critical review of the literature is required. Modern sociotechnology is the basis for the developments [3], as well as the definition of key concepts: events, in order to replace the tayloristic task as the criteria for work division, and communication (in Habermas sense, reworked by Zarifian [4]) for the coordination of work. The aim is the design of work groups not predefined – their configuration depends on the events to face. So, they are reconfigurable groups, open, with variable geometry – variable composition, variable part of the process attended etc. Contrary to many work group or team usual implementation, groups have to be thought as part of the organisation: they must be linked to the whole firm or company. Processes design and direct production
groups integration with staff and white collar activities is key. The methodology proposes the following design rules: 1. Discussion, formalisation and divulgation of design principles / declaration of values of the project 2. Definition of operational processes 3. Parallelisation (flow design) 4. Segmentation (division of work among groups) 5. Criteria for choosing process technology (equipments, software etc.) 6. Information systems, production of information, spaces for communication 7. Control system to face to events 8. Support social systems: wage system, careers, training, symbolism Field research was conducted mainly in Brazil but also in France (automotive, paper, food) and Italy (chemicals, automotive, machinery). 2
LITERATURE REVIEW ON ORGANISATIONAL DESIGN TO FLEXIBILITY The traditional tayloristic model of work organization has received many criticisms due to its rigidity and inadequacy to current competitiveness linked to product innovation, time to market in not commoditised products etc. Sociotechnical systems design, born at Tavistock Institute, London, and well spread into the US and North Europe, was one of the first post-II World War critics to taylorism, as well as an attempt to create a formal approach towards flexibility and better work conditions. A production system would be composed of two subsystems, the social and the technical, that must be joint optimized. Cherns [5], and Gerwin and Kolodny [6] have systematized the approach in principles, highlighting: variance control – as closest as possible to its source; multiskilling; low specification on how the work should be done but high specification on what is expected from the group; work participation and good work conditions. The latter authors, based on Louis Davis works [2], propose an organisational design based on the analysis of production variances, the establishment of technical borders among organizational unities, and
work (job) design. In the approach, flexibility is treated as the ability of a given group to cope with variances. Although not explicit, the approach lead to groups with fixed composition and fixed part of the process to work on. Many psychologists of work developed the idea of individual self-belonging into a group - a given one [7], [8]. But the approach lacks of considering practical design rules – some of the original authors from Tavistock Institute were even against such rules. It gives few tools for design, and there is the trend for practitioners to adopt classical (tayloristic) rules, that are simple, logic, written – even if they could be not recommendable to the type of company here in focus. The so-called lean approach also lacks design rules [9]. Although many ideas of the lean approach can be useful in any kind of company, actually it is much more focused on the rationalisation of mass production operations (reduction of waste etc.) than on the creation of value by product innovation in a turbulent environment. “Toyotalike” teams are closed groups, with fixed composition, working on a given area of the process, strictly guided by a supervisor. Mintzberg’s configurations [10] are quite powerful to analyse organisations, but does not make easier the design of an organization structure. Adhocracy is the closer configuration to our goals; the idea of self adjustment among persons as a way of coordination is perhaps the most important contribution of the author. Rarely an organizational paper puts emphasis on coordination, like Mintzberg do; normally the central issue being division of work whether for prescription or critics. But self adjustment must be build, designed. Sitter et al. [3] have gone further in design methods improving socio-technical tradition. They introduced the so-called modern sociotechnology, by stating there are no singles social and technological subsystems; instead, subsystems would be functions like production, design etc. The aim of organisational design would be to grant the maximum of control of the system, in the sense of the capability to drive it to desirable states. This is an important goal to cope with unpredictable situations. The spirit of the approach is the more the ease to change work organisation, the more flexible the organisation. 3
DEPICTING CLASSICAL STRENGTH TO BUILD OPTIONS In order to build coherent and pertinent alternatives to the classical functional (or tayloristic) approach to organisation, there is a need to build new concepts to replace the existent ones, as well as systematic design rules to make possible for practitioners and academics to employ them, improve them, and replace them. Briefly, the strength of classic approach is simplicity and adequacy between the physical process of production and the financial process of producing profits. Taylor’s task concept means a given work method – rationalised human motion, in a given workplace, both designed to minimize the (standard) time. By reducing standard time of the operations management reduces cost; in an environment that buys all the products produced it means that low production costs leads to higher profits. So, the reduction of time standards is directly linked to profitability in a stable business – predictable market, no challenged from competitors, no substitutive products, no alternative technologies etc. For most of the businesses, this time has gone. Task is the unity of classical work organisation. It is the criteria for work division. Coordination is embedded in technical stuff (like assembly lines) or performed by hierarchy (supervision / management). To build an
alternative means to substitute task concept as the criteria for division of labour and coordination by technical means or hierarchy. 4
EVENTS AND COMMUNICATION AS THE CRITERIA FOR WORK NUCLEATION AND COORDINATION
4.1 Event replaces task Veltz and Zarifian [11] have introduced the concepts. Event is an unpredictable, unusual, random but important situation that must be faced by a given production system in a given moment. It is not a physical phenomenon, it is a social and economical phenomenon – for instance, the fall of a dry branch of a tree is an event for the gardener but not for those who watch it by the window. An event is not physical in the sense that it can never “happens” – for instance, an event is the procedures developed by a work team in a packing process of a food industry in order to keep the line running by adjusting the guides of the machine that would be un-calibrated without the intervention: the action prevented the physical happening but it was an event in the sense that was judged important by the team and have mobilised competences to face it. An event is singular, not foreseen in norms that describe or evaluate situations. It is unpredictable, making irrelevant the imposition of time and motion to workers because, due to the event, prevails a time logics different from the “regular” course of production It can not be reduced to a fact of the objective world since are the members of the social world that makes an event from an occasion – if currently computers breakdown are more important than in the 60s, this does not mean that there are more breakdowns with existing computers than with the valve ones; however, current economic conditions for profitability and strategies of time to market put light on breakdowns; and an event is inherent to a given situation. In that sense, event conceptually replaces the classical notion of task as the key concept for work division. Around events competences are mobilised to face them; workers are mobilised – better, mobilise themselves – to cope with them. A given breakdown can require a group of workers able to deal with mechanics, electronics, pneumatics; another one, only a mechanical. Competences required in a given moment depend on the particular event, being unpredictable. So, a rigid division of labour like in traditional or closed socio-technical group (or self managed groups, semi-autonomous groups, self-directed teams – there are many names for the same scheme), with the same workers with the same competences every shift, every day, every week, is not the most efficient choice. 4.2 Zarifian’s concept of communication to replace classical coordination The question is how to set mobile groups configuration – in other words, groups with variable geometry, open. In order to workers mobilize required competences to face an event it lacks a crucial point: how to coordinate in between workers? How would workers deal with tactical questions regarding the decision criteria in a given moment? For instance, there are alternatives in a breakdown: a) to fix up the equipment as soon as possible because the client has urgency to receive the goods; b) to promote an in-depth diagnosis of the equipment in order to prevent further problems, in order to elevate production system reliability, since in two days a new order will be produced and the pace will be increased, without much time for machinery adjustment. Both alternatives are acceptable depending on the given situation. The crucial question is coordination – classically, the supervisor defines the priority but it uses to take a long time between the breakdown and the information of the priority for the workers.
19th International Conference on Production Research
This idle time would be reduced if workers could discuss tactical issues, validating priorities of the production system linked to the specific business “game” being played. Veltz and Zarifian [11], and Zarifian [4], proposed the concept of communication as the central one to replace the classical notion of coordination. Communication is a polysemic word; in management it is quite common to use it as information, orders (in hierarchical meaning), companies’ letters or news, information systems (e-mails) etc. But information in the sense of saying to others what to do is not a useful notion for the paper’s purposes. Communication must be understood as mutual intercomprehension among people, as proposed by Zarifian. It is different from information, orders etc. And there are three dimensions, as can be depicted from the examples in this section: a) the cognitive one – which competences to mobilise?; b) the normative one – which rules, which priorities to guide the intervention?; c) the expressive one, concerning instigation and rewards (wages, careers etc.) The cognitive dimension is linked to a mutual comprehension and understanding on which competences – and which professionals – are required to face up a given event. For instance, one technician in electronics must consider important the mobilisation of pneumatic competences in order to keep the system running. Moreover, the technician must recognize in his colleague the competence and the aptitude to perform that job: cognitive dimension is not only technical but also social. 5
DESIGN RULES
5.1 Discussion, formalisation and divulgation of design principles / declaration of values of the project To make explicit values and design principles is of uppermost importance to disseminate decision criteria among design team – many designers tend to participate in specific parts of the project; written values help to keep team homogeneous, limiting time consuming discussions on which are the basis of the project. Moreover, it makes clear the compromises assumed by management towards employees, what is quite important in redesign of existing organisations.
other line, cell etc. If a new product or packing must be tested and production calibrated to it, one single part of the production process is occupied, the other ones keeping the normal production running. This rule tries to isolate and to circumscribe the effect of variations, perturbations, events. It is crucial for the need to cope with unpredictable goals. Sure, an investment analysis must be made in each case. But it must be considered the effective performance of the whole production system in medium-long term instead of the usual static analysis, which uses to consider nominal performance of the equipment in a stand-alone situation. It is quite common that a decision to buy two machines instead of one with nominal double capacity pays off. 5.4 Segmentation (division of work among groups) Refers to criteria for the division of work, what sociotechnical traditional approach names as borders allocation of a group. As far as it is impossible for a single group to take care of a whole complex production system, the goal is to design macro-groups that can divide themselves depending on each single situation, each single event. Moreover, exceptionally this macro group can even ask for extra-workers from another macro-group as it happens in the case discussed below. As proposed initially, event took the place of task as the criteria for work mobilisation and work division. Moreover, according to a whole set of authors and to our own researches, the main problems of efficiency in a more integrated production system have origins because of the boundaries in-between organisational unities, not inside them. It is useless to optimise a time standard of a workplace if subsequent work in process waits hours to be transferred – there are lots of other examples concerning relations between production sectors, with maintenance, scheduling, The following criteria for the basic initial allocation of macro-groups are proposed: a) to internalise organisational borders in the group; b) the cycle of analysis, treatment and action due to an event must be preferably allocated to a single group; c) secondary criteria, according to the case: 3T – technology homogeneity, territory (space), time (shifts etc.), and numbers of people in a macro-group.
5.2 Definition of operational processes Definition of main processes linked to strategic axles. Operational processes are transversal to the company, not functional / departmental. Four rules are proposed for operational processes design: a) definition of firm’s strategic axles; b) discussion of the processes linked to each axle associated with a joint discussion of all processes; c) analysis of each process focusing on its characteristics and criteria of performance; d) Identification of the activities of each process, and discussion with social actors (workers, supervision etc.) on the sequence of activities, weakness, strengths and improvement opportunities, as well as indicators for each activity to guide quotidian action.
5.5 Criteria for choosing process technology (equipments, software etc.) Process technology – machinery and software – uses to have an imbedded organisational and work division logic, setting up restrictions for the design of the organisation. For instance, an automated system to control a continuous flow process offered to a company had terminals for quality control department (there is none) with special passwords inaccessible to production workers. Firstly, the company decides the firm’s strategy and the organisational logic, macro-design and structure, and only after elects process technology. Technology must be compatible with strategy and structure considered in a horizon of time.
5.3 Parallelisation (flow design) Puts light on the analysis of production flows aiming at make the system more flexible and reliable to cope with external uncertainties (mainly linked to market and regulation) and internal ones (linked to failures in equipments, the need to test new products in line, lack of inventories, lack of delivery from suppliers etc). The general approach is to prefer modular production systems, small lines in parallel instead of a long line, cells instead of functional sector, concurrent engineering etc. Parallelisation means that there are options for the flow: if a machine is out of order, production can be redirected to
5.6 Information systems, production of information, spaces for communication Information systems must be developed after parallelisation and segmentation, never before. They must be coherent to organization logics and goals. Some spaces to promote communication (in the sense of mutual understanding among people) shall be designed. Formal meetings to discuss the future are essential to promote the normative communication dimension, like the one to discuss production scheduling involving production workers, maintenance and sales. Meetings to discuss the past are essential both to improve professional skills and
the cognitive dimension of communication, like those to analyse the actions performed to cope with events in the past. Communication is a social construction, not a technical affair. 5.7 Control system to face to events (drivability) Since the generic goal of the type of production systems in focus is the capability to reach unpredictable and changing goals instead of a predefined bunch of goals, system’s drivability (or control) is key. Drivability derivates from the structure, depends on how organisational structure was designed and implemented. Drivability cycle is composed by a) perception / representation of the state of the process by workers; b) judgment on what to do; c) action. So, if the approach to design organisational structure is top-down, from strategy to groups, the approach to drivability / control systems is bottom-up. The objective is to eliminate control needs by allocating them in lower organisational levels. Information systems discussed in the precedent item must be designed to help the representation of the state of the process and to make control cycle and actions to face events easier. 5.8 Support social systems: wage system, careers, training, symbolism It is crucial to design wage and career system compatible with organisation demands. Traditional over-divided wage scheme induces in the opposite direction instead of incentivizing self mobilisation of workers, independent of their background, to face an event. Management by competence is a good starting point; social or organisational skills should be considered as well as the technical ones. It helps a lot in workers’ engagement, in expressive dimension of communication. Nobody works for nothing. Workers evaluation is also crucial. Evaluation criteria and management acts are the most important signalisation on what is only a discourse, and what is real in everyday life at work. 6 A WELL SUCCED CASE A Brazilian chemical consumer goods plant shows an impressive case of reconfigurable organisation. It has three mini-plants defined by different processes: chemical batch flow to produce the main raw material; physical transformation of raw materials – heat, pressure, gravity etc.; packing. There are three macro-groups per shift, one in each process, only two hierarchical levels (workers and management). Processes 1 and 2 run 24 hours/day. Process 2 is vertical, 30 meters high, with several stages, but a single group run it – around 30 persons / shift. Workers talk by phone and radio; every one can ask the help of others to face an event. There are a control room where computers terminals are installed but every member of the group can go there. Who is in charge of the computers obviously has more information and a broader view of what is going on. Shifts were designed to overlap 30 minutes. During first 15’ there is a meeting with representatives of the groups ending job with all workers of the next shift for process 2 and 3; the last 15 minutes there are informal conversations on the job. One day, the group heading process 1 faced a huge problem; they decided that the best would be to stop the process. But this would mean stopping deliveries. It was 3 a.m. They called up process 2 to investigate their level of inventory; they researched on external delivery schedule of their product, called up an external client asking for their level of inventories and if it would be possible to postpone delivery for one day. After that, they stopped the process. At 9 a.m. the manager arrived, got panic because the miniplant was down, but relaxed after being informed of the procedures taken in the precedent shift.
Evaluation depends on general performance (in order not to dislocate variances or problems to other group or shift), and to specific measures of state of every mini-plant. The plant has broken all the performance records of the group – they have plants all over the world, becoming benchmarking. 7 ACKNOWLEDGMENTS The author is gratefully thanked to Marcelo Aulicino (Polytechnic School and Petrobras), Roberto Marx (polytechnic School), Luiz Felipe Cortoni (LCZ Consulting) and people from the companies researched. CNPq – Brazilian Council for Science and Technology financed partially the research. 8 REFERENCES [1] Emery F.E., Trist E.L., Socio technical systems. In: Emery F.E., ed. Systems thinking. Harmondsworth, Penguin, 1969, 281-296. [2] Davis L., Taylor J.C., 1972, eds., Design of jobs, Harmondsworth, Penguin, 1972. [3] Sitter L.U., Dankbaar B., Hertog J.F., From complex organizations with simple jobs to simple organizations with complex jobs, 1997, Human Relations, 50-5, 497-534. [4] Zarifian P., Travail et communication: essai sociologique sur le travail dans la grande entreprise industrielle. Paris, PUF, 1996. [5] Cherns A., Using the social sciences. London, Routledge & Kegan Paul, 1979, 310-340. [6] Gerwin D., Kolodny H., Management of advanced manufacturing technology: strategy, organization and innovation, New York, J. Wiley & Sons, 1992. [7] Tjosvold D., Team organization: an enduring competitive advantage. Chichester, J.Wiley & Sons, 1991. [8] Katzenbach J. R., Smith D. K., The discipline of teams, 1993, HBR, mar./apr., 111-120. [9] Woomack J.P., Jones D.T.; ROOS D., The machine that changed the world, New York, Macmillan, 1990. [10] Mintzberg H., Structures in fives: designing effective organizations, Englewood Cliffs, Prentice Hall, 1993. [11] Veltz P., Zarifian P., Vers de nouveaux modèles d'organisation? 1993, Sociologie du Travail, XXXV-1, 3-25.
