ARTICLE IN PRESS

Int. J. Production Economics 122 (2009) 419–428

Contents lists available at ScienceDirect

Int. J. Production Economics

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

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.128, 05508-900 São Paulo—SP, Brazil

a r t i c l e in fo abstract

Available online 21 June 2009 The paper proposes a methodology and design rules for organisational structures facing
Keywords: higher necessity of rapidly reconfigure themselves to cope with unpredictable
Reconfigurable organisation situations—new markets, new products, changing mix of production, problems in
Open groups production process or flows etc. It implies changing and often conflictive criteria for
Organisational design production goals and for the allocation of work. The methodology was developed based
Flexibility on a large field action research and consulting. Their basis is the design of auto-
Efficiency reconfigurable working groups—or groups with variable geometry, depending on the
events to face.
& 2009 Elsevier B.V. All rights reserved.

1. Introduction moreover an ability to face to unpredictable situations.

The need to have a high degree of equipment utilisation to
Contemporary operations efficiency must be closed make the system economically viable means to identify,
linked to companies’ strategy. Many businesses are run in act and solve problems very fast, and to act accordingly to
unpredictable and turbulent environments—new compe- the strategy pursue at that moment; moreover, means to
titors, new technologies, new products, changing financial act to avoid the emerging of dysfunctions, disruptions,
conditions etc. Frequently it means that a production quality problems, flow problems etc.
system must be designed and implemented without a Clearly, production structure must be designed to cope
clear vision of the range or mix of products it will be asked with contemporary needs of flexibility. Organisational
to deliver. Rigid organisational schemes do not fit well to literature is rich in discussing flexible schemes compared
this situation. In a caricatural picture, the reduction of to the traditional tayloristic/fordist ones. The main
time standards of single operations does not lead to approach in that sense, the so-called socio-technical
increases in profits as it happens in a growing and systems (Emery and Trist, 1969; Davis and Taylor, 1972),
predictable market, a ‘‘buying’’ market. Decision rules is a real advance but it has some conceptual problems
for production scheduling can change according to the (Sitter et al., 1997; Eijnatten and Zwaan, 1998) and mainly
situation—not only cost, sometimes product performance, it lacks a tooling kit to orient organisational design—the
sometimes time to market, sometimes a combination of design rules.
them. Automation, low inventories and outsourcing This paper proposes a methodological approach and
complicate the scene: integrated production systems are design rules for organisational design for companies that
more suitable to export varieties or problems to other face competitive and technological environments like the
workplaces, sections, departments, plants or companies. one discussed before. To do so, a brief critical review of the
Automated processes, with remote control and guidance literature is required. Modern socio-technology (Sitter
by direct workers, require an abstract ability to be run, et al., 1997) is the basic framework for our development,
as well as some key concepts: events, in order to replace
the tayloristic task as the criteria for work division, and
 Tel.: +55 11 30915363x484; fax: +55 11 30915399. communication (in Habermas sense, reworked by Zarifian,
E-mail address: [email protected] 1996) for the coordination of work. The aim is the design

0925-5273/$ - see front matter & 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.ijpe.2009.06.015
 ARTICLE IN PRESS
420 M.S. Salerno / Int. J. Production Economics 122 (2009) 419–428

of work groups, not predefined—their configuration, their Table 1

geometry, their boundaries depends on the events to face. Profile of the companies in the field research.
So, they are reconfigurable groups, open, with variable
Company Product/process Research tools/activities
geometry—variable composition, variable part of the
process attended, variable goals to be achieved etc. Brazil
Contrary to many work group or team usual implementa- Pack1 Filling/packaging 1,2,3,4,5,7,9,10,11,12,13,14,15,16,17
tion, groups have to be thought as part of the organisa- process
Pack2 Filling/packaging 1,2,3,4,5,10,13,14,15
tion: they must be linked to the whole firm or company. It process
is insufficient to think of the design only at the direct Pack3 Filling/packaging 1,2,3,4,5,7,9,10,13,16
production level. A more integral design should be process
comprehensive, thinking of the whole system—engineer- Food1 Food 1,2,3,4,5,7,9,10,13,14,15,16,17
Food2 Food 1,2,3,4,5,10,13,14,15,16
ing, management, indirect work in general. Processes
Food3 Food 1,2,3,4,5,6,7,10,13,14,16
design and direct production groups integration with staff Food4 Food 1,2,3,4,5,9,10,13,14,16
and white collar activities is of upmost importance for an Food5 Food (ice cream, 2 1,2,3,4,5,7,9,12,13,15
integral design. plants)
The methodology proposes the following design rules: Auto1 Commercial vehicles 1,3,4,5,6,8,9,11,12,13,14,15,16,17
Auto2 Passenger vehicles (2 1,3,5,6,9,11,12,13,14,15,17
plants)
1. Discussion, formalisation and divulgation of design Auto3 Passenger vehicles (3 1,2,3,4,5,6,8,9,10,11,12,13,14,15,16,17
principles/declaration of values of the project plants)
Auto4 Passenger vehicles 1,3,4,6,8,11,12,13,14,15,16,17
2. Definition of operational processes
Auto5 Passenger vehicles (3 1,2,3,4,5,6,8,10,11,12,13,14,15
3. Parallelisation (flow design) plants, headquarters)
4. Segmentation (basic division of work among groups) Auto6 Commercial vehicles 1,2,3,4,5,6,8,11,12,13,14,15
5. Criteria for choosing process technology (equipments, (plant, headquarters)
software etc.) OfficeStuff1 Office materials 3,4,9,13,14,17
Machin1 Machinery & 4,6,8,9,11,12,13,14,15
6. Information systems, production of information, precision mechanics
spaces for communication PersonalCare1 Personal care 1,2,3,4,5,7,9,10,12,13,14,15,16
7. Control system to face to events (chemical)
8. Support social systems: wage system, careers, training, PersonalCare Personal care 1,3,4,13,14
2 (chemical)
and symbolism.
PersonalCare Personal care 1,2,3,4,5,13,14,16
3 (chemical)
Chemical1 Chemical 1,2,3,4,5,7,9,10,11,12,13,14,15,16
2. Methodological issues Chemical2 Chemical 1,2,3,4,5,12,13,14,16
Chemical3 Chemical (2 plants) 1,3,4,5,11,12,13,14
The proposal was developed based on a large field Chemical4 Chemical 1,2,3,4,5,9,10,13,14,16
Chemical5 Chemical (oil 1,3,5,6,8,10,11,12,13,14,
research and some consulting during late 90s’ and refinery, 3 plants)
beginning of the 2000s. It mixes aspects of survey, case Aero1 Aeronautical 1,2,3,5,6,10,11,12,13
studies and in-depth action research. It was conducted
France
mainly in Brazil but also in France (automotive, paper, FRAuto1 Vehicles (3 plants; 1,3,4,5,6,13,14,16,17
food industries) and Italy (chemicals, automotive, ma- headquarters/R&D
chinery). It was discussed and applied in organisations centre)
like Unilever (food and consumer goods), Rhône-Poulenc FRAuto2 Vehicles 1,3,4,5
FRFood/96 Food 1,2,3,4,5,13,14
(chemical), Danone (food), Rockwell (auto parts/mechan- FRCardboard Corrugated 1,2,3,4,5,9,10,13,14
ical), Petrobras (oil refinery and engineering), Embraer cardboard & boxes
(aircrafts), R&D sectors of some companies as well as
Italy
some attempt in public sector, and others. Some visits in ITComp Auto components 1,2,3,4,5,6,9,10,12,13,14
plants in other countries served to think of the proposals, ITAuto Vehicles 1,3,6,9,11,12,13,14
but were not formally conducted with methodological ITMachin Machine tools 1,4,6,9,11,12,13,14
care; these visits happened in Germany, South Korea, ITElect Electrical stuff 1,2,3,5,12,13,14

Uruguay, Belarus, Russia, where, apart from Uruguay and Germany

some companies in Germany, there were a language DEAuto1 Vehicles (3 plants) 1,3,4,5,6,8,10,11,12,13,14,16
barrier and time shortage to perform the methodology.
(1) Direct observation; (2) interviews with main company manager; (3)
Some companies in Brazil, France and Italy also fit in this interviews with plant managers; (4) interviews with human resources
latter category. manager; (5) interviews with other managers and technical staff; (6)
In all companies there were an in-depth discussion interviews with trade union leaders; (7) group dynamics with direct
with managers, plant visits, the following of events and workers performed in the company; (8) group dynamics with direct
workers performed in the Union; (9) discussions with direct and indirect
the analysis of how workers faced it, focusing on the ease workers; (10) interviews with bosses (supervisors etc.); (11) discussions
or the barriers the organisational structure put to the with plant committees/union representatives in the company; (12)
perception, diagnosis and action to face events. These discussions with trade union technical staff; (13) analysis of company’s
interviews with managers and technicians have two materials; (14) analysis of secondary materials (paper, statistical data
etc.); (15) supervision of PHd, MSc, or graduation theses; (16) interviews
characteristics: (a) apart from a general overview of the
with managers in the headquarters; (17) interviews in plants in the
company, the production and its challenges, they were home country (and vice-versa, in Brazil).
 ARTICLE IN PRESS
M.S. Salerno / Int. J. Production Economics 122 (2009) 419–428 421

based on an open guideline focused on the way the ment. Socio-technical systems design, born at Tavistock
organisation uses to perceive, judge and act on problems/ Institute, London, and well spread into the US and North
events; (b) they were performed at first in a formal way Europe, was one of the first post-II World War criticisms
(in the office), but also during plant visits, lunch time, to taylorism, as well as an attempt to create a formal
management meetings etc. approach towards flexibility and better work conditions. A
In most companies there were discussions with work- production system would be composed of two subsys-
ers on the same focus discussed above; in few but very tems, the social and the technical, that must be joint
relevant cases there were group dynamics with workers, optimized. Cherns (1979), and Gerwin and Kolodny (1992)
dynamics with managers, long periods of production have systematized the approach in principles, highlight-
observation and discussions and the following of work ing: variance control—as closest as possible to its source;
activities in the workplaces. Table 1 shows the profile of multiskilling of the workforce; low specification on how
the companies researched and the activities performed in the work should be done but high specification on what is
each one. expected from the work group; work participation and
These dynamics are not an easy tool, they depend on good work conditions. The latter authors, based on Davis
the accordance of the companies; it means that some and Taylor (1972), proposed an organisational design
workers will be absent of the production for some hours based on the analysis of production variances that would
or will be in extra-hours. But they are a very powerful tool guide the establishment of technical borders among
to comprehend the real functioning of the work, of the organisational unities, and work (job) design. In the
organisation, its problems and potentials. One dynamic approach, flexibility is treated as the ability of a given
was only performed after some plant visits and detailed group to cope with variances. Although not explicit, the
observation of the production, and after interviews with approach lead to groups with fixed composition and fixed
managers and technicians to get the strategy, production part of the process to work on. Many psychologists of
goals regarding the strategy and the market, organisa- work developed the idea of individual self-belonging into
tional practices, changes that happened, resistances, a group-a given one (Tjosvold, 1991; Katzenbach and
future vision. We performed separate dynamics with Smith, 1993), supporting the idea of a fixed group.
direct workers and indirect production ones (mainte- But the approach lacks in considering practical design
nance, quality staff etc.). After an introduction explaining rules—some of the original authors from Tavistock
our research activity, we made a deal on time period of the Institute were even against such rules. It delivers few
activity, confidentiality etc. Then, the following exercise tools for design, it is not well formatted for practitioners;
was proposed: the workers received a large sheet of paper, this lead to a the trend for practitioners to adopt classical
some magazines, scissor, and glue. The aim was to made (tayloristic) rules, that are simple, logic, written—even if
two panels, each one representing one situation of work they could be not recommendable to the type of company
(before and after a change, a production problem, an here in focus. The so-called lean approach (Womack et al.,
special event—it depended on each case, on the previous 1990) also lacks design rules for the kind of flexibility we
data collected). A discussion was conducted based on the are considering here. Although many ideas of the lean
panel ‘‘readings’’, guided by a roadmap previously pre- approach can be useful in any kind of company, actually it
pared, in order to focus on the events, on the problems, on is much more focused on the rationalisation of mass
the way the production deals with the events. Each production operations (reduction of waste etc.) than on
dynamic took around 4 h. the creation of value by product innovation in a turbulent
In some companies (Pack1, Food1, PersonalCare1, environment. ‘‘Toyota-like’’ teams are closed groups, with
Chemical1) we performed dynamics also with managers. fixed composition, working on a given area of the process,
In all companies that we performed in-depth research strictly guided by a supervisor.
(action research), with dynamics, we made several meet- Mintzberg’s configurations are quite powerful to analyse
ings with management to discuss our findings and organisations (Mintzberg, 1993), but does not make easier the
proposals. design of an organisational structure. Adhocracy is the closer
In the cases where the dynamics took place at the configuration to our goals; the idea of self adjustment among
Union, the techniques utilized were: (a) verbalization (the people as a way of coordination is perhaps the most
workers were asked to describe their activities in some important contribution of the author. Rarely an organisational
key events); (b) drawings of the production process and paper puts emphasis on coordination, like Mintzberg did;
workplace, with flows, to anchor the discussion. normally the central issue is division of work whether for
prescription or its criticism. But self adjustment must be built,
designed and implemented—that is the challenge of a
3. Literature review on organisational design to practitioner that could be simplified by a formal theoretical
flexibility and methodological approach with appropriate design rules.
Literature on virtual enterprise focuses mainly in task
The traditional tayloristic model of work organisation allocation and managing the system; Martinez et al (2001)
has received many criticisms due to its rigidity and discuss the concept of VE and cite the question of non-
inadequacy to current competitiveness linked to product hierarchic and distributed control but do not develop it.
innovation, time to market in not commoditised products Hülsmann et al. (2008), concerning supply chain manage-
etc. And, as pointed out by Veen-Dorks (2005), traditional ment, proposes the adoption of autonomous cooperation
management control is not aligned with such environ- and control, characterised by autonomy, decentralized
 ARTICLE IN PRESS
422 M.S. Salerno / Int. J. Production Economics 122 (2009) 419–428

decision-making, interaction, heterarchy, and non-deter- unpredictable, unusual, random but important situation
minism, to achieve flexibility in global supply chain that must be faced by a given production system in a given
operations, but they also do not discuss design rules. moment. It is not a physical phenomenon, it is a social and
Sitter et al. (1997) have gone further in design methods economical phenomenon—for instance, the fall of a dry
improving socio-technical tradition. They introduced the branch of a tree is an event for the gardener but not for
so-called modern socio-technology, by stating there are those who watch it by the window. An event is not
no singles social and technological subsystems; instead, physical in the sense that it can never ‘‘happens’’—for
subsystems would be functions like production, product instance, an event is the procedure taken by a work team
design, etc., each one joining both social and technical in a packing process of a food industry (Food1) in order to
aspects—actually, it is not possible to isolate the keep the line running by adjusting the guides of the
‘‘technical’’ from the ‘‘social’’ and vice-versa. The aim of machine that would be un-calibrated without the inter-
organisational design would be to grant the maximum of vention. The action prevented the physical happening of
control of the system, in the sense of the capability to packing problems; it can be classified as an event in the
drive it to desirable states. This is an important goal to sense that was judged important by the team and have
cope with unpredictable situations. The spirit of the mobilised competences to face it. It was random and
approach is: the simpler the organisational structure, unpredictable because nobody could know when and how
the more complex the work, the easier the control of the to calibrate the system; it was unusual in the sense that it
system to cope with variability, the more flexible the does not fit in the predetermined routine.
organisation. An event is singular, not foreseen in norms or rules that
describe or evaluate situations. It is unpredictable, making
4. Depicting classical design rules strength to build irrelevant the imposition of time and motion to workers
options because, due to the event, prevails a time logics different
from the ‘‘regular’’ course of production. It cannot be
reduced to a fact of the objective world since are the
In order to build coherent and pertinent alternatives to
members of the social world that make an event from an
the classical functional (or tayloristic) approach to
occasion—if currently computers breakdown are more
organisation, there is a need to build new concepts to
important than in the 60s, this does not mean that there
replace the existent ones, as well as systematic design
are more breakdowns with existing computers than with
rules to make possible for practitioners and academics to
the valve ones; however, current economic conditions for
employ them, improve them, and further on replace them.
profitability and strategies of time to market put light on
Briefly, the strength of the classic approach is simpli-
breakdowns; and an event is inherent to a given situation.
city and adequacy between the physical process of
In that sense, event conceptually replaces the classical
production and the financial process of producing profits.
notion of task as the key concept for work division, or, in
Taylor’s task concept means a given work method—ratio-
better words, as the key concept to organise people in a
nalised human motion, in a given workplace, both
group, to set a group, an open, variable group. Around
designed to minimize the (standard) time. By reducing
events competences are mobilised to face them; workers
standard time of the operations management reduces
are mobilised—better, mobilise themselves—to cope with
cost; in an environment that buys all the products
them. A given breakdown can require a group of workers
produced, it means that low production costs leads to
able to deal with mechanics, electronics, pneumatics;
higher profits. So, the reduction of time standards is
another one, only a mechanical. Competences required in
directly linked to profitability in a stable business—pre-
a given moment depend on the particular event, being
dictable market, no challenged from competitors, no
unpredictable. So, a rigid division of labour like in
substitutive products, no alternative technologies etc.
traditional or closed socio-technical group (or self man-
For most of the businesses, this time has gone.
aged groups, semi-autonomous groups, self-directed
Task (In Taylor’s sense, 1911) is the unity of classical
teams—there are many names for roughly the same
work organisation. It is the criteria for work division.
scheme), with the same workers with the same compe-
Coordination is embedded in technical stuff (like assem-
tences every shift, every day, every week, is not the most
bly lines) or performed by hierarchy (supervision/man-
efficient choice.
agement) or technocracy (methods or planning office). To
Traditional and modern socio-technical and design
build an alternative means to substitute task concept as
approaches conceptualise variability or variances analysis
the criteria for division of labour and coordination by
(Gerwin and Kolodny, 1992; Cherns, 1979; Sitter et al.,
technical means or hierarchy/technocracy.
1997). We consider the concept of event superior of the
variability one. Variability means a deviation from a
5. Events and communication as the criteria for work standard, from an average; it is not well adapted for
nucleation and coordination systems with variable goals, where multiple standards
often change, requiring a trade off among them. And with
5.1. Event replaces task as the main concept for work the concept of event it is possible to analyse and to set
division/auto-organisation design rules also for situations where a variation was not
produced, but the goals of the moment, the configuration
Veltz and Zarifian (1993) have introduced the concept of the production system, of its indicators, mean the need
of event in the organisational discussion. Event is an for an action.
 ARTICLE IN PRESS
M.S. Salerno / Int. J. Production Economics 122 (2009) 419–428 423

5.2. Zarifian’s concept of communication replaces the diagnosis of the failure, changing components etc.); (c) to
classical way of work coordination use the breakdown as an opportunity to learn, to train
people, to discuss with engineering improvements to get
The question is how to set a mobile group config- better productivity or quality etc. Each possibility has its
uration—in other words, groups with variable geometry, own logic, is acceptable depending on the situation,
open. In order to workers mobilise required competences depending on the needs of the whole production system,
to face an event it lacks a crucial point: how to coordinate of the tactics of the company in that moment. If there is a
in between workers? How would workers deal with crucial order to be delivered in the same day, maybe the
tactical questions regarding the decision criteria in a best would be to fix the machinery as soon as possible; if
given moment? For instance, there are alternatives in a the machinery is producing a part of an order that will be
breakdown, like: (a) to fix up the equipment as soon as delivered in some days, the need for the machinery to
possible because the client has urgency to receive the keep running well in the near future could lead to the
goods; (b) to promote an in-depth diagnosis of the second possibility; and if there is no such pressure of time,
equipment in order to prevent further problems, in order it would be misleading not to take advantage of the
to elevate production system reliability, since in two days moment for improvements, learning, training, preparing a
a new order will be produced and the pace will be future best performance.
increased, without much time for machinery adjustment. In order for an open group to decide which possibility
Both alternatives are acceptable depending on the given would be the best, they have to know and, moreover, to
situation. The crucial question is coordination—classically, validate, the norms and the tactic issues of production
the supervisor defines the priority but it uses to take a plan and sales being played at the moment. We will see a
long time between the breakdown and the information of case on it below. A key question for the company and for
the priority to the workers. This idle time would be the organisational design is how to promote communica-
reduced if workers could discuss tactical issues and tion in the normative dimension, making possible for
validate priorities of the production system linked to the people that have gathered themselves to face an event
specific business ‘‘game’’ being played. (according to the cognitive dimension) to decide which
alternative to follow in their action. To have access to
Veltz and Zarifian (1993), and Zarifian (1996), inspired information on production scheduling and planning, sales
by Habermas (1987), proposed the concept of commu- priorities, inventories etc. is crucial for the decision. To
nication as the central one to replace the classical notion participate in the decision of monthly/weekly/daily plan-
of coordination. Communication is a polysemic word; in ning could be of uppermost importance for workers to
management it is quite common to use it as information, develop normative mutual inter-comprehension, but only
flows, orders (in hierarchical meaning), companies’ letters a real participation, with voice, not only with ears, will
or news, information systems (computer networks, e- lead to it: if a worker says that the machine will not
mails) etc., as Koufteros et al. (2007) have pointed out. But support the charge, and the management charges it
information in the sense of saying to others what to do is anyway, there is no development of communication, and
not a useful notion for the paper’s purposes. Communica- workers tend to abandon the meetings or to be there in a
tion must be understood as mutual inter-comprehension bureaucratic sense (obliged by the bosses).
among people, as proposed by Zarifian. It is different from And, of sure, nobody would involve themselves in a
information, orders etc. And communication in that sense scheme like the one being proposed here without a formal
has three dimensions, as can be depicted from the retribution, since responsibility grows, productivity
examples in this section: (a) the cognitive one—which grows, managers are free to deal with more strategic
competences to mobilise? (b) the normative one—which issues—this is an important question regarding workers
rules, which priorities to guide the intervention?; (c) the autonomy, often misleading.
expressive one, concerning instigation and rewards Based on the original framework of Sitter et al. (1997)
(wages, careers etc.). and the concepts of event and communication, we will
The cognitive dimension is linked to a mutual propose design rules for an organisation dealing with
comprehension and understanding on which competen- unpredictable goals.
ces—and which professionals—are required to face up a
given event. For instance, one technician in electronics
must consider important the mobilisation of pneumatic 6. Design rules for organisations dealing with
competences in order to keep the system running. More- unpredictable goals
over, the technician must recognize in his colleague the
competence and the aptitude to perform that job: 6.1. Discussion, formalisation and divulgation of design
cognitive dimension is not only technical but also social. principles/declaration of values of the project
The normative dimension is crucial. It guides the way
to approach a problem, to face an event. Taking the To make explicit values and design principles is of
example of machinery breakdown cited in the beginning uppermost importance to disseminate decision criteria
of this section, there are some alternatives to face the among design team—many designers tend to participate
event: (a) to minimize downtime in the short time (fix it in specific parts of the project; written values help to keep
in the minimum gap of time); (b) to minimize downtown team homogeneous, limiting time-consuming discussions
in the medium time (performing a more comprehensive on what is the basis of the project. Moreover, it makes
 ARTICLE IN PRESS
424 M.S. Salerno / Int. J. Production Economics 122 (2009) 419–428

clear the compromises assumed by management towards variations, perturbations, events. It is crucial for the need
employees, what is quite important in redesign of existing to cope with unpredictable goals.
organisations. Sure, an investment analysis must be made in each
case. But it must be considered as the effective perfor-
mance of the whole production system in medium-long
6.2. Definition of operational processes term instead of the usual static analysis, which uses to
consider nominal performance of the equipment in a
Definition of main processes linked to strategic axles. stand-alone situation. One of the main benefits of
Operational processes are transversal to the company, not parallelisation concerning investment analysis is the
functional/departmental. They can be defined as a possibility to modularise investment, postponing part of
cooperation of activities for the concretisation of a global it. For instance, the company invests in one line, is able to
objective, directed to a final client that is the same for all deliver product to the market, and, according to sales
activities. Each process should have: (a) a global perfor- growth, it can invest in other parallel lines—this strategy
mance indicator that formalises its main goal (e.g., a time is being used by Petrobras in a new oil refinery being
to market, a cost); (b) an organisation that links their designed, showing that even in continuous flow, where
activities; (c) a co-responsibility of the actors towards the process technology usually puts important constraints
global objective, and regarding single activities they are to organisational design, there is some room to paralle-
involved. The global performance objective should be lisation.
deployed to the activities, but remains the main reference By the other hand, it is quite common that a decision to
for the evaluation of the process. Locally, in key activities, buy two machines instead of one with nominal double
there are indicators of means, not of objectives. Activities capacity pays off. In Food5 we have seen the opposi-
can be the basis for an ABC/M—activity based costing/ te—the introduction of a complex packing machine
management. replaced two existing one, but scrap has increased (due
Four rules are proposed for operational processes to higher speed and higher complexity to adjust) and
design: (a) definition of firm’s strategic axles; (b) discus- flexibility to test new products has decreased, what is
sion of the processes linked to each axle associated with a critical in that business. In DEAuto1, automation in a
joint discussion of all processes; (c) analysis of each 1.5 km synchronous assembly line had contributed to
process focusing on its characteristics and criteria of inefficiency; if a single screw was not in the position in a
performance; (d) identification of the activities of each station, the whole line was down, and workers have to
process, and discussion with social actors (workers, walk long distances to access the special gates to go into
supervision etc.) on the sequence of activities, weakness, the line. The effect is a loss of up to 20% of working time,
strengths and improvement opportunities, as well as an extra area for rework, and 5.500 workers instead of the
indicators for each activity to guide quotidian action. 3.400 initially planned (including shifts, coordinators,
The better way is to involve employees in the discussion rework etc.). This is an extreme case of the problems of
to set and to rethink processes; it is a way of learning and a non-parallel flow design; DEAuto1 has modified this
of professional and organisational development (Galbraith conception in subsequent lines.
and Lawler III, 1995). In the auto industry there used to be an idealisation of
the Toyota scheme of stopping the assembly line—the
hidden issue was the ease to run overtime in Japan; in
6.3. Parallelisation (flow design) new Toyota plants, long lines have been broken down in
parts, separated by a buffer designed to cope with small
Puts light on the analysis of production flows aiming at stops in each segment, without stopping the whole line.
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 6.4. Segmentation: division of work among groups or
equipments, the need to test new products in line, lack allocation of group boundaries
of inventories, lack of delivery from suppliers etc). The
general approach is to prefer modular production systems, Refers to criteria for the division of work, what socio-
small lines in parallel instead of a long line, cells instead of technical traditional approach names as borders alloca-
functional sector, concurrent engineering etc. To some tion of a group. As far as it is impossible for a single group
extent, the discussion on modular design and supply (as to take care of a whole complex production system, the
modularisation in auto industry) has a similar logic of proposal is to design macro-groups that can divide
parallelisation—economies of investment, parallelisation themselves depending on each single situation, each
of assemblies etc. (Salerno, 2001). single event, like in Chemical1, Food1, Pack1 and others.
Parallelisation means that there are options for the Moreover, this macro group can even ask for extra-
flow: if a machine is out of order, production can be workers from another macro-group as it happens in the
redirected to other line, cell etc. If a new product or case discussed below.
packing must be tested and production calibrated to it, As proposed initially, event took the place of task as the
one single part of the production process is occupied, the criteria for work mobilisation and work division. More-
other ones keeping the normal production running. This over, according to a whole set of authors and to our own
rule tries to isolate and to circumscribe the effect of research, the main problems of efficiency in a more
 ARTICLE IN PRESS
M.S. Salerno / Int. J. Production Economics 122 (2009) 419–428 425

integrated production system are due to boundaries in- of filling system etc. In order to permit auxiliary workers
between organisational unities, not inside them. It is to participate, formal training was provided; the end of
useless to optimise a time standard on a workplace if the lines was adapted to allow a small buffer of finished
subsequent work in process waits hours to be transferred, products before packaging in the boxes, to make possible
or that product will not be sold anymore—there are lots of for workers to move along the lines to regulate the filling,
other examples concerning relations between production talk to people at the plastic blow in the moment etc. As
sectors, production with maintenance, scheduling or result, machine occupation increased and quality pro-
product design, quality problems etc. blems were reduced to 10% of the previous level.
The following criteria for the basic initial allocation of
macro-groups are proposed:
6.5. Criteria for choosing process technology (equipments,
software etc.)
(a) to internalise organisational borders in the group;
(b) the cycle of analysis, treatment and action due to an Process technology—machinery and software—uses to
event must be preferably allocated to a single group; have imbedded organisational and work division logics,
(c) one activity defined during the discussion/design of setting up restrictions for the design of the organisational
operational processes must be under the responsi- structure and dynamics. For instance, an automated
bility of a single group. system to control a continuous flow process offered to
(d) Secondary criteria, according to the case: 3T—tech- Chemical1 had terminals for a quality control department
nology homogeneity, territory (space), time (shifts with special passwords inaccessible to production work-
etc.), and numbers of people in a macro-group. ers. As the company has firstly define their organisational
principles and concepts (like in Section 6.1), and decide do
not segment apart production and quality control tests, do
Fig. 1 shows a real case, the discussion of segmentation
not have a QC department, not to have direct supervisors,
in Pack1, a plant that produces liquid detergents for
the supplier was asked to change the parameters of the
manual dishwashing. There are several filling/packaging
system. Without the formal procedures on organisational
lines; each one receives plastic bottles produced in-house
design, Chemical1 would have bought a supervisor level, a
and plastic caps and labels from suppliers. The lines must
quality control sector etc.
be calibrated for filling the bottles, closing with caps,
So, the company must firstly decide the firm’s strategy
pasting labels; at the end, the bottles are arranged in
and the organisational logic, macro-design and structure,
corrugated cardboard boxes and arranged to go to ware-
and only after elects detailed process technology. Tech-
house. The plant faced several quality problems—bottles
nology must be compatible with strategy and structure
with less volume that it should have causing problems
considered in a horizon of time. Of course, there are some
with final consumers, supermarkets and governmental
business strictly linked to a process technology, like oil
agencies; wrong labels; bottles discarded due to imperfect
refinery, petrochemical and others; it limits the degrees of
closing, and so on. As consequence, machine occupation
freedom for organisational design (like in parallelisation)
was low, leading to inefficiency and higher costs.
but the example of Petrobras shows that there are room
The work was divided traditionally: one fixed machine
for parallelisation and, at least, it is always possible to
operator for each line, the only person that regulated the
discuss how to configure production control software,
line; 5–7 auxiliary workers in the beginning and at the
passwords, terminals allocation etc., like in Chemical2,
end of the line to guarantee the right position of the
Chemical3, PersonalCare3. This point leads us to discuss
plastic bottles and to move filled bottles into the boxes.
information systems.
The company started analysing the introduction of
groups; the inspiration was the case of PersonalCare1,
which was successful in improving quality and productiv- 6.6. Information systems, production of information, spaces
ity years before. A dynamic conducted with workers for communication
showed they had a very clear notion on where the
problems came from, mostly from plastic bottles; the Information systems must be developed after paralle-
problem of weight was due to foam formation during lisation and segmentation, never before. They must be
filling; etc. But they had no time to discuss with the coherent to organisation logics and goals. Some spaces to
machine operator due to the fixed position—if one gets promote communication (in the sense of mutual under-
out of the line the bottles could fall down on the standing among people) shall be designed. Formal meet-
ground—and the distance; they only notice someone ings to discuss the future are essential to promote the
when the operator or the supervisor went to the final part normative communication dimension, like the ones to
of the line. discuss production scheduling involving production work-
The company studied three design options, as showed ers, maintenance, sales, general management, and the
in Fig. 1 and decider to introduce the open group participation of direct workers in teams responsible to
approach. The same group is responsible for plastic blow design expansions, new lines, technological changes etc.
(production of the bottles), for filling, packing and delivery At Food1, the dynamics we realised with workers
to the warehouse, internalising previous borders. The idea showed that they participate in a weekly meeting for the
is to keep under control of a single group the main events planning of the next week but their opinion were often
to face and action to perform—regulation of plastic blow, disregarded. As a result, their participation became
 ARTICLE IN PRESS
426 M.S. Salerno / Int. J. Production Economics 122 (2009) 419–428

P
Process overview Option1: fixed group only with the former
sl s machine operators
B
a o
o l Plastic blow lines to supply
s p Aims at lowering work costs by flexible allocation of machine
o
r filling/packaging lines
pt w operators along the different lines. The option was discarded:
i o Pack1 has realised that manual packing workers were most
cr sensitive to quality problems since they manipulate the bottles. And
Packing lines: some with manual it creates new borders, tending to elevate the period of time to treat
feeding and packaging events that requires the intervention of more than one unity

filling / packaging line

operation
area
filling / packaging line

operation
filling / packaging line
area

Option 2: a fixed group by each line Option 3: one macro-group considering plastic blow
and final packaging
Each group is homogeneous (same jobs) and decides members’
allocation along the line. The option seeks to increase efficiency
due to mutual cooperation. Problem: the need to intensive training The same group is responsible for plastic blow, filling and
for ex-non operators (unskilled). packaging. Although plastic blow and packing are hundreds of
meters far from each other, with a wall in between, the idea is to
improve efficiency through immediate action
b b when there is a problem due to plastic bottles
l l in the packing line – an analysis showed that
filling / packaging line o o this is the main problem of line stopping and
w w product quality. The group could rapidly
identify a problem, reallocating workers to face
it. After three months of experience and
training, quality problems were reduced by
80%; productivity grew substantially.

filling / packaging line

filling / packaging lines

Fig. 1. Segmentation options at Pack1.

formal, an obligation, and they stopped to point out meetings like that in some companies; they can be really
problems in machines, raw materials etc.; the efficiency effective if conducted in a positive way, to learn, not to
declined (measured by machine occupation), delivery find guilty. Communication is a social construction, not a
problems arose due to the difficulty to achieve scheduling technical affair.
goals. During a meeting with management we presented
the diagnosis; the general manager decided to accept any
suggestions by the workers in order to regain trust; in the 6.7. Control system to face to events: controllability,
next meeting, he asked several times on the problems drivability
with equipment and others. The workers told about a PLC
(programmable controller) with a problem not yet fixed Since the generic goal of the type of production system
although in the maintenance schedule; this problem in focus here is the capability to reach unpredictable and
would lead to a lower level of production. The manager changing goals instead of a predefined bunch of goals,
accepted to change the planning, negotiated with sales; system’s drivability (or controllability) is crucial. Driva-
sales contacted the main client (the major Brazilian bility derivates from the structure; depends on how
supermarket network) to redefine delivery; it was not a organisational structure was designed and implemented.
problem for the client. The company used the gap of time Drivability cycle is composed by (a) perception/represen-
gained with the changing in production scheduling to fix tation of the state of the process by workers; (b) judgment
the PLC—a simple but decisive action to regain efficiency on what to do; (c) action. So, if the approach to design
and, moreover, to regain trust in order to rebuild the organisational structure is top-down, from strategy to
conditions for normative communication. groups, the approach to drivability/control systems is
Meetings to discuss the past are essential both to bottom-up. The objective is to eliminate control needs by
improve professional skills and the cognitive dimension of allocating them in lower organisational levels. Informa-
communication, like those to analyse the actions per- tion systems discussed in the precedent item must be
formed to cope with events in the past. We followed some designed to help the representation of the state of the
 ARTICLE IN PRESS
M.S. Salerno / Int. J. Production Economics 122 (2009) 419–428 427

process and to make control cycle and actions to face and 3; in the last 15 min there are informal conversations
events easier. on the job.
One night, the group heading Process 1 faced a
huge problem that could leave to a breakdown in the
6.8. Social support systems: wage system, careers, training,
reactor—it had broken before. The question was whether
symbolism
to stop production or whether to continue at least until
daylight and office hour. They decided that the best would
It is crucial to design wage and career system
be to stop the process. But this would mean stopping
compatible with organisation demands. Traditional over-
deliveries for some time. It was 3 am. They called up
divided wage schemes induce in the opposite direction
Process 2 to investigate their level of inventory; they
instead of incentivizing self mobilisation of workers,
researched on external delivery schedule of their product,
independent of their background, to face an event.
called up an external client asking for their level of
Management by competence (Zarifian, 2001) is a good
inventories and if it would be possible to postpone
starting point; social or organisational skills should be
delivery for one day. After that, they stopped the process.
considered as well as the technical ones. It helps a lot in
At 9 am the manager arrived, got panic because the
workers’ engagement, in the expressive dimension of
mini-plant was down, but relaxed after being informed
communication. Nobody works for nothing. Workers
of the procedures taken in the precedent shift.
evaluation is also crucial. Evaluation criteria and manage-
There are no quality or maintenance sectors. When a
ment acts are the most important signalisation on what
machine axle went out of order, a worker called up a
only a discourse is, and what is real in everyday life at
maintenance company that had a contract with the
work. In the same way, to have exclusive restaurants,
company, informing the problem, called up a taxi and
parking area, cash dispensers for managers do not fit well
went to the service supplier with the axel to machine a
with a company with few hierarchical levels that depends
new one—without asking permission, without informing
on workers’ self mobilisation and responsibility: if a
the management. When a leak happened in Process 2,
worker can not have lunch in the same place of a manager,
they asked directly, by phone, for workers of Process 3
why should him act as a manager?
(packing) to help them.
Chemical1 is an extreme and very well succeed case of
7. A well succeed case: an open organisation, the reconfigurable organisation. It seems as if there is one
worker–manager single group of workers that reallocate themselves
according to production needs. Managers let the
A Brazilian chemical consumer goods plant (Chemi- production run and have more time to concentrate on
cal1) shows an impressive case of reconfigurable organi- strategic or tactical issues—how to improve efficiency,
sation. It has three mini-plants defined by different how to improve the business, how to keep the system
processes: chemical batch flow to produce the main raw running well.
material; physical transformation of raw materials—by The plant has broken all the performance records of the
heat, pressure, gravity etc.; packing. There are three whole company (efficiency, quality, cost and the lowest
macro-groups per shift, one in each process, only two period of time since the inauguration to achieve the
hierarchical levels (workers and management); there is a expected standards, which were surpassed in the
large part of skilled workers; the company seeks for dual sequence), which has plants all over the world, becoming
skill (like mechanical and electrician, electrician and benchmarking. This made visible an important problem.
electronics etc.) and pays for the formation. The company had a global wage classification, and the
Processes 1 and 2 run 24 h/day. Process 3 utilises ten blue collars of the plant were at the top. The headquarters
parallel packing lines; there is a kind of reference group by did not accept a change in the system to fit this plant,
line, but these groups are open, can send and receive other although all its success. So, other companies in the region
people according to the events to face—for instance, if the start offering more to the workers, and some have
packing machine adjusted for one priority product has a changed company. When it was starting to became a
problem in the final stage (which puts several packs of the common fact, the general manager of the business at
product into a larger box in order to send it to the Chemical1 ignored the headquarters and moved most of
warehouse or directly to the distribution truck), people the main skilled blue collar workers to a white collar
from other lines can help packing manually part of the classification. Management also must be flexible.
production. Process 2 is vertical, 30 m high, with several
stages, but a single group runs it—around 30 persons/
shift. Workers talk by phone and radio; every one can ask 8. Conclusions
the help of others to face an event. In Process 2 there are a
control room where computer terminals are installed but The paper proposed a framework, a methodology and
every member of the group can go there. Who is in charge rules for the design of organisations. The goal is to build
of the computers at the moment obviously has more an organisation able to reconfigure itself according to the
information and a broader view of what is going on. Shifts challenges it faces in order to cope with unpredictable
were designed to overlap 30 min. During first 150 there is a goals. To do so, at first a new framework was build, based
meeting with representatives of the groups ending on the concepts of event and communication as the
shift with all workers of the next shift for Processess 2 mutual inter-comprehension among people, respectively,
 ARTICLE IN PRESS
428 M.S. Salerno / Int. J. Production Economics 122 (2009) 419–428

to replace the classical concepts of task (time standard, (Polytechnic School), Luiz Felipe Cortoni (LCZ Consulting,
motion standard, predefined workplace) and coordination who proposed the method of the dynamics) and people
by hierarchy, technocracy or imbedded in technical stuff. from the companies researched. Brazilian Science and
Event is the concept for self mobilisation of workers; Technology Research Council (CNPq/PQ) financed partially
communication is the coordination tool, conceived based the research.
on three aspects: the cognitive, the normative and the
expressive. In that sense, it is possible to act according to References
the challenges to face, to changing goals.
But it is not enough to conceptualise. Design rules Cherns, A., 1979. Using the Social Sciences. Routledge & Kegan Paul,
must be outlined to support the designers with a formal London, 310–340.
Davis, L., Taylor, J.C. (Eds.), 1972. Design of Jobs. Penquin, Harmonds-
tool. Based on modern socio-technology, modified with
worth.
the concepts of event and communication and the scheme Eijnatten, F.M., Zwaan, A.H., 1998. The Dutch IOR approach to
of open, reconfigurable groups, design rules were though organizational design. Human Relations 51 (3), 289–318.
to build an organisation in a systemic (or integral) way. Emery, F.E., Trist, E.L., 1969. Socio technical systems. In: Emery, F.E. (Ed.),
Systems Thinking. Penquin, Harmondsworth, pp. 281–296.
The rules are comprehensive, delivering a framework that Galbraith, J.R., Lawler III, E.E., (Orgs.), 1995. Organizando para competir
guides a corporation throughout the process of thinking of no futuro: estratégia para gerenciar o futuro das organizac-ões. São
the relation strategy-structure and the use of internal Paulo, Makron (Chapter 4).
Gerwin, D., Kolodny, H., 1992. Management of Advanced Manufacturing
resources to take advantage in changing conditions of Technology: Strategy, Organization and Innovation. Wiley, New York.
competition, in innovation. Habermans, J., 1987. Théorie de l’agir communicationnel. Fayard, Paris.
Implementing the proposed organisational scheme Hülsmann, M., Grapp, J., Ying, L., 2008. Strategic adaptivity in global
supply chains: competitive advantage by autonomous cooperation.
requires clear values, management dedication, cohesion International Journal of Production Economics 114 (1), 14–26.
and firmness to keep the path. Although many criticisms Katzenbach, J.R., Smith, D.K., 1993. The discipline of teams, Harvard
and a lot of literature, companies are accustomed to Business Review, March/April, 111–120.
Koufteros, X.A., Nahm, A.Y., Cheng, T.C.E., Lai, K.-H., 2007. An empirical
classical structures and to hierarchical coordination. It is assessment of a nomological network of organizational design
not so taken for granted that a plant which supposedly constructs: from culture to structure to pull production to
should have high flexibility goes beyond ‘‘Japanese style’’ performance. International Journal of Production Economics 106,
468–492.
quality tools and a suggestion system. It is not usual the
Martinez, M.T., Fouletier, P., Park, K.H., Favrel, J., 2001. Virtual enterprise:
discussion with process engineers and with information organisation, evolution and control. International Journal of
technology professionals on the requirements each sys- Production Economics 74, 225–238.
tem should deliver to fit organisational needs—normally, Mintzberg, H., 1993. Structures in Fives: Designing Effective Organiza-
tions. Prentice Hall, Englewood Cliffs, NJ.
the discussion turns around internal efficiency of the Salerno, M.S., 2001. The characteristics and the role of modularity
software or the equipment, but real efficiency is the one in the automotive business. International Journal of Automotive
achieved in day-to-day operation, not in a stand alone Technology and Management 1 (1), 92–107.
Sitter, L.U., Dankbaar, B., Hertog, J.F., 1997. From complex organizations
equipment users’ manual. An equipment very efficiency with simple jobs to simple organizations with complex jobs. Human
does not mean production effectiveness/efficacy regarding Relations 50 (5), 497–534.
company’s goals. Real efficiency and efficacy is achieved Taylor, F.W., 1911. The Principles of Scientific Management. Harper Books,
New York.
by a collective understanding among management and all Tjosvold, D., 1991. Team Organization: An Enduring Competitive
types of workers, their mutual comprehension of the Advantage. Wiley, Chichester.
goals, and the way they are organised, or organise Veen-Dorks, P., 2005. Management control and the production
environment: a review. International Journal of Production
themselves, to act with equipments, information systems,
Economics 93–94, 263–272.
suppliers and customers. Veltz, P., Zarifian, P., 1993. Vers de nouveaux modèles d’organisation?
Sociologie du Travail XXXV (1), 3–25.
Womack, J.P., Jones, D.T., Roos, D., 1990. The Machine that Changed the
Acknowledgments World. Macmillan, New York.
Zarifian, P., 2001. Objetivo competência: por uma nova lógica. Atlas, São
Paulo.
The author is gratefully thanked to Marcelo Aulicino Zarifian, P., 1996. Travail et communication: essai sociologique sur le
(Polytechnic School and Petrobras), Roberto Marx travail dans la grande entreprise industrielle. PUF, Paris.