logistics everywhere
TRANSCRIPT
-
8/12/2019 Logistics Everywhere
1/11
Since it began to be used by certain
large US manufacturing firms after
World War II, logistics management
has become increasingly effective.
After the initial attempts which met
with varying degrees of success,
both the definition and applicationsof logistics management now seem
to be quite settled: Logistics refersto the systematic management of thevarious activities required to movebenefits from their point ofproduction to the customerTheconcept of benefits is a multifacetedone that goes beyond the product orservi ce itself to include issues regardi ng
tim ing, quantity, suppor ting services,location and cost. So a basic definiti on
of logi stics is the conti nuous process
of meeting customer needs by ensurin gthe availabil ity of the right benefitsfor the ri ght customer, in the quantity
and condition required by thatcustomer, at the tim e and pl ace thecustomer wants them (Gourdin,2001:2). Our aim is therefore not to
try to question the traditional
approach of logistics management,
but rather to expound the thesis
that it is possible to significantly
enlarge the scope of logistics
management in order to include
other fields. These comprise some(and probably many) issues that
have an important bearing on the
management sciences as a whole.
More precisely, we claim that there
are several examples of disciplines
which could each be reformulated,
at least in part, and which should
be reexamined as problems in
logistics management. In this first
paper devoted to expounding thisthesis, we focus for the most part
on two aspects of management
sciences chosen essentially
at random: advertising and
knowledge management. These
disciplines, or domains of activity,
are usually considered to be
quite far removed from logistics as
well as from each other. However,
we claim that the basic problems
in these disciplines can be
represented as instances of a quite
basic and general formulation oflogistics management, where one
can focus even more than usual on
the basic concepts of optimization
and flows. Defending such a thesis
requires two components: one
analytic and the other synthetic.
For the whole approach to make
sense, the analysis needs to reduce
logistics management to its very
core, and this core needs to be
expressed in a way that is universal
enough for it to be detected as an
essential part of the make-up ofother disciplines. At this very
early stage of our thinking, the
synthetic part is limited to the
It is now an accepted f act th at logistics man ag ement a ims to provide
service qua lity, cost reduction an d eff icient a ft er-sales suppo rt th roug h
an eff ective monito ring o f supply chain operat ions. In order to achieve
these goals, logisticians have learned to define clearly the problems they
have to ma nag e in terms of product f low s optimization and t o use
metho ds that a re both rigorous and rational in order to solve them.
In this pap er, we pro pose tha t it shou ld be possible to a pply and to
generalize the logistical approach to other fields, including those as
improba ble as advertising and know ledge mana gement
Logistics Everyw here:Can a Logistical Approach
help Tackle the Majorityof Managerial Problems?
36Supply Chain Forum An International Journal Vol. 4 - N1 - 2003 www.supplychain-forum.com
Charles Tresser
M athematical Sciences Department
IBM Thomas J. Watson Research C enter, USA
tresser@ us.ibm.com
Gilles Pach
Centre de Recherche sur le Transport
et la Logistique
Universit de la M diterrane
A ix M arseille II, France
pache@ iut.univ-aix.fr
-
8/12/2019 Logistics Everywhere
2/11
37Supply Chain Forum An International Journal Vol.4 - N1 - 2003 www.supplychain-forum.com
case study of the two examplesthat have been previouslymentioned.
This paper examines thefoundations of logistics management,
starting with the original aimsand continuing up to the
most recent approaches andstrategies (including supply chainprocess integration). More than
just reviewing the subject, we willexplain our belief that, despite its
obvious evolution and the progressthat has been made, logisticsmanagement has not changed in a
fundamental way, and has notundergone any radical change sinceit began to be considered as
belonging to the managementsciences. A typical MBA course
on logistics management wouldp ro fes s t o i n c reas e t hemanufacturing firms profitability,
t h u s r es p o n d i n g t o t heexpectations of the shareholdersbut this motivation hides the basic
thinking behind it. In other words,l o gi s t i c s ( s up pl y c hai n )
management, which today is stilllimited to playing a supporting rolein helping to improve
manufacturing and selling, is inessence organized around the
concepts of optimization and flows.One must also identify theobjectives that are to be optimized
and under what constraints suchan optimization must be performed(or approximated). We would like
to emphasize that the twodisciplines which are taken as
examples have been chosen atrandom. We simply wanted oneunusual example to discuss briefly
and one example which we could
discuss at greater length at thisearly stage of the research and
which would provoke a debate. Asthis paper is interdisciplinary, we
have made an effort to spell out alltechnical acronyms, even thosethat are standard in logistics
management. 1
An or thodox v iewof lo gistics
Logistics management, which hasbeen subjected to a slow
maturation process, firstly in theUSA and then in Europe, is deeply
rooted in the military sciences, thetransition happening when thesoldiers who had been working on
its development as a modernscience came back to civilian
business life after World War II.Until quite recently, it was foundedon product flows optimization, a
perspective that has beenchallenged by the new trendstowards supply chain process
integration advanced by someresearchers. We take the view that
this challenge is not motivated by
any serious breakdown in the role
of logistics management, as itsevolution over recent years has
confirmed the presence of onemajor objective: to optimize
inbound and outbound logisticaloperations in order to increase thefirms profitability and thus
respond to the shareholdersexpectations. In fact, thecommonalities of the problem
solving methods of logisticsmanagement are so broad that not
only do they apply to both recentand historical trends, but they alsoapply to fields quite distant from
product flows optimization.
Product flow s optim ization
Historians have recently shown
that supplying cities with food fromdistant lands in order to avoid
famine is a problem with a clearlogistical nature, which arose inEurope at least as far back as theMiddle Ages (see, for instance,
Abad [2002]). Should we take this,therefore, as the first step towards
the development of a logisticalapproach? Certainly not, as it isimpossible to identify any real
global thinking aiming towards thebetterment of the means ofdistribution of provisions: mankind
was at that time most oftenpowerless in the face of random
factors over which it had nocontrol. So, while it is certain thatlogistical problems existed, and to
some extent were routinely solvedlocally through necessity, the veryidea of logistics managementwas
still to be born. It appears muchlater and its origins are
undoubtedly military. Anyencyclopedia would refer to itprimarily as the combination of
transportation, provisioning, andaccommodation designed to ensurethe success of troops in combat. If
detailed enough, the encyclopediawould probably add that the father
of military logistics is theNapoleonic General Antoine-Henri
Jomini, the author of the famousPrcis de lart de la guerre, ounouveau tableau analytique des
principales combinaisons de la
stratgie, de la grande tactique et dela pol i t ique mi l i ta i re, published in
1838, in which he presents logistics
1. After a first version of the paper had been
circulated, we received very encouragingcomments from Alan Hoffman, Bill Grey,
Nicholas Zissos, and from an anonymous
reviewer of Supply Chain Forum: An
Internati onal Journal. We also received useful
informati on from Naoki Abe, John Forrest, and
John Tomlin about v ario us uses of the methods
of logistics in aspects of advertising operati ons.
What these colleagues, now all from IBMsResearch Center, hav e taught us is that several
authors have, for instance, used optim izati on
techniques to make a better choice in the
distribution of advertisements, in classical
settings and mor e recently on the World Wide
Web. However, it seems that here we are
postulating a new approach, whi ch in the case
of advertising aims to reformulate the
discipline in a holistic way as a problem of
logistics management. On the other hand,
these comments given to us on the appli cationsof optimi zation methods in advertising lead us
to expect that the approach developed here
should bear frui t at some stage in the future and
not just remai n a philosophical perspective.
When considering
disciplines where the
methods of logistics
(supply chain)
management have not
previously been
applied, it is
necessary to identifythe entities that need
to flow, to where they
must flow and to
what benefit.
-
8/12/2019 Logistics Everywhere
3/11
as being one of the six parts of the
art of war (Jomini, 2001).Throughout the twentieth century,
the field of military logisticsunderwent spectacular developments,
culminating in Operation Overlord,the landing of the Allied troops on
the occupied European Continent
on June 6, 1944. The French
coastline was fortified and
defended by particularly well-equipped German troops, therefore
it was essential to plan the different
stages of the invasion meticulously
in order to avoid a repeat of the
cruel failure of Operation J ubilee,
which was attempted on the
beaches of Dieppe in March 1942.
To this effect, the Anglo-American
Joint Command assembled theorganizational and logistical means
necessary to anticipate and
overcome the constraints of the
terrain. More recently, during
Operation Desert Storm in 1991, the
media referred frequently to the
importance of the role played
by logistics as an essential
complement to the firepower of theAllied armies in the successful
outcome of the military operation.
What is more, the quality of
logistics improved dramaticallyduring the 2003 Iraq war. In a
recent study for the Institute of
Comparative Strategy, Richardot
(2002) also referred to the crucially
important role played today by theCombat Logistics Force (CLEF) in
the organization and efficiency of
the US Navy. Therefore, the link
between logistics and the success
of military operations remains
particularly strong; even if today it
is widely acknowledged that
logistics is also (and mainly?) a
part of management. We mustremember, however, that the
application of logistics in
manufacturing firms just after
World War II was a direct result of
what had been learned from
Operation Overlord and the overall
process of winning the war. The
distances which had to be covered
in the liberation of Europe obligedthe formalization of re-supply and
transport processes, which were
recognized as being applicable to
American firms needing to supplycustomers located hundreds or
thousands of miles from the
factories.
This explains why logistics
management was first perceived
by the American Marketing
Association (1948) as the planningof the operations linked to the
dispatch of goods from the factoryto the consumer (transport,
handling and warehousing). The
goal was to adapt and apply the
tools from Operational Research,
tools that had initially been
conceptualized in a military
context. We will refrain fromdelving further into the history of
logistics management (see for
instance Lambert et al. [1998]), but
we want to stress that its military
origins strongly influenced its
development until the 1970s.
During that time span, the mainquestions continued to revolve
around the implementation ofconstrained optimization methods
to help in the dispatching of goods
from their origination points to
their destinations. One of the best
known examples of such a problem
is the design of the delivery round
that minimizes the total distance
covered as a function of the
stores to be served in a givengeographical area. This is an
example of the Traveling SalesmanProblem, simple to formulate but
which quickly becomes too difficult
for any computer to solve when
the number of stores increases!
Logistics management was thus
limited to the one best way in
terms of product flows monitoring.
This focus was justified, as markets
that were previously local and
national became progressivelyglobal for multinational corporations,
a fact which accentuated the
problem of dispatching goods atlow cost and with high service
quality. As well informed observers
of logistics management will know,
Magee (1968) made a significant
step forward by introducing the
concept of logistical processes and
the three systems:(1) the procurement system;
(2) the operating system; and
(3) the physical distribution system.
However, we believe that a
continuity exists between theapproaches of the American
Marketing Association (1948) andof Magee (1968), as the analysis
remains based upon product
flows optimization. Magees (1968)fundamentally important contribution
is to stop limiting the role
of logistics management to
that of a physical distributionsystem. Contradicting the formerreductionistviews, he argued that
the manufacturing firm must also
optimize product flows inside the
production cycle (work inprogress [WIP] management) and,
downstream to the operating
system, inside the physical supply
cycle (materials management), if it
wants the total logistics system or
pipeline to function correctly. But
is it possible to view this as a truly
radical departure when, for Magee
(1968), the issue remains theoptimum planning of logistical
operations using increasingly
sophisticated tools, such as
program evaluation and review
techniques (PERT) and material
requirement planning (MRP)? By
the time his book appeared, the
world had grown more complexand the markets where goods were
consumed had become larger and
more distant, creating the need
for an even more effective
coordination of transport, productionand warehousing activities. More
than ever before, logistics
management experienced an urgent
need for managerial proceduresthat would use mathematical
algorithms to bring about the best
solution at the right moment.
In fact, even a few years later,
and despite its revolutionary
contributions, the way Heskett
( 1977) en vi s i o ns l o gi s t i c smanagement remains captive to the
same product flows optimizationparadigm. The economic landscape
has changed since the beginning
of the recession in the Western
countries, and in the particularly
difficult context of price competition,
firms must adjust their logistics (in
the physical distribution system,the operating system and the
procurement system) to adhere as
closely as possible to the
downstream demand in order to
reduce costs. Long before such
reasoning became fashionable,Heskett (1977) indirectly introduced
the idea that efficiency dependedon a double paradigm: lean
38Supply Chain Forum An International Journal Vol.4 - N1 - 2003 www.supplychain-forum.com
-
8/12/2019 Logistics Everywhere
4/11
production and agility. From aconcrete standpoint, it is the actualsales, and no longer the sales
forecast, that must trigger theoperations of the supply of raw
materials, production, transport,etc. Thus, Heskett (1977) implicitlyconceptualized the just-in-time
(J IT) strategy, at the very sametime as it first began to be appliedin Japan by Toyota! But while that
vision was incontestably aninnovative one, it remained
centered on manufacturing firms,whose survival depended on theoptimization of their own logistics
systems. A good JIT strategy willfirst induce a reduction in the stocklevels of goods and materials, while
it may also be the case that stocklevels abruptly increase at the
retailer or wholesaler locations.
Supply chain processintegration
Do the recent trends in logisticalthinking challenge the orthodox
vision in terms of flowsoptimization? It is possible toanswer yesand noat the same
time. According to the majority of
authors (see, for example,Underhill [1996], and Mentzer et al.[2000]), all of the firms involved invertical exchange relationships
insist on the importance of the newintegrative dimension of thelogistical processes: specifically,
that in order to bring the rightproduct at the right time and at the
right place to the consumer, it isnecessary to coordinate all of thelogistical operations along the
supply chain. Effective logisticsmanagement thus becomes the
means to generate a close anddurable collaboration betweensuppliers, manufacturers, retailers,
and third party logistics (TPL)service providers, so that togetherthey create value for the customer
(Grey et al., 2003). However, wemust also admit that the aim of asupply chain process integration
remains to create value for theshareholders, which in turn
requires enforcing product flowsoptimization.
At a fundamental level, even if thisvision is sometimes disputed, asupply chain corresponds to aset of operations supporting: (1)
the logistical activities of the
procurement of materials in anetwork of first-tier and second-tierproviders; (2) the transformation of
materials into semi-finished orfinished goods; and (3) the physicaldistribution of the finished goods
to customers. The basic idea isthat these activities and operations
must be perfectly synchronizedwithin the framework of supplychain management (SCM), so that
customers can benefit from thebest possible service quality at the
lowest possible cost. Stevens(1989) suggested the existence offour steps leading to the emergence
of genuine SCM: Step 1 is distinguished by
complete functional independence
in so far as each functionoperates quite separately from
all the others, protected bybricks (stocks) allowing fordifferences between their
operating rhythms. In step 2, manufacturing firms
recognize the urgent need for aminimum coordination betweenrelated functions, e.g. marketing
and physical distribution, so asto eliminate a number ofmalfunctions.
Next, step 3 is there to improvethe process of coordination by
implementing the comprehensive
planning of internal flows fromdownstream to upstream,
leading to: Step 4, when supply chain
members finally become aware
that they are merely part of awhole.
The virtue of this evolutionaryapproach is to highlight the fact
that SCM refers to a succession of
trade-offs (Lambert et al., 1998).Trade-offs occur within a firm (e.g.,
between purchasing and materialsmanagement, materials management
and production, etc.). The purposeis to find overall solutions, which,through intra and inter-organizational
collaboration, will avoid a wasteof resources and will increasethe profitability of supply chain
members. An interesting exampleis that of convenience goods
manufacturers, where largeretailers and TPL service providerswork together to create the
packaging for a new product toreduce logistical costs from thefactory to the store shelves. One
could also quote the example of anautomotive manufacturer and itsmain suppliers deciding to
implement a common milk-rounddelivery system: a collection of
vehicles that visit suppliers in apredetermined and negotiatedorder, at pre-agreed times during
the night, from which there followsa reduction of congestion at theassembly plant (Lamming, 1983).
I n h i n d s i gh t , an d w h i l e
acknowledging the progress madein logistics management over thelast fifteen years or so up until now
in 2003, it is clearly appears thatSCM does not represent a radicaldeparture. Of course, SCM
emphasizes the importance of theindispensable collaboration
between supply chain members inan extended enterprise (Naccaratoet al., 2000), but the object of
analysis still remains the product and information flows. On
the other hand, thanks totechnological progress, andnotably computer technologies
(from core hardware to softwareand algorithms, and more recently,pervasive computing, which
considerably enhances reach andcommunication), it is now much
easier to optimize under constraintthan it was, for example, fifteenyears ago, and this is where the
essential change resides. Forinstance, a small and medium-sizedenterprise (SME) can now in real
time tackle certain optimizationproblems that would have been too
complex for most large companiesin the 1980s. Thanks to these newtechnologies, the supply chain
members can adapt faster to all
sorts of environmental changes andexecute effective time-based
strategies, where the object is toconceive new products faster than
the competition and then tomanufacture and deliver them tothe customer more quickly and
with no mistakes. For over half acentury, logistics management hasthus proven its capacity to improve
the performance of manufacturingfirms and, on a more global level, of
networks of firms, by relying on alogistical approach with a universalambition. From this point of view,
might logistics managementbecome the means to help resolvemanagerial problems in areas far
39Supply Chain Forum An International Journal Vol.4 - N1 - 2003 www.supplychain-forum.com
-
8/12/2019 Logistics Everywhere
5/11
removed from product flowsoptimization?
The central t hesis anda f irst e xample
Clearly, the goal we have set toourselves is quite ambitious and
challenging, but the rewards couldbe enormous for most of themanagement sciences. The process
began with a strong hunch thatlogistics was extremely prevalent,
with possible examples to be foundin most of the sciences, from thePrinciple of Least Action in physics,
to the way a cell works in biology.We will refrain from formulating ourthesis on such general grounds and
instead, we will concentrate beforehand on a formulation that focuses
on the management sciences. Theformulation we provide is still quitebrief but we expect that many
people will join us in the comingadventure, and will help us todevelop the fundamentals as well
as the applications of the ambitiousprogram implicitly outlined in our
thesis. We have randomly picked adiscipline from within or close tothe field of management that would
be as improbable a subject aspossible to use, namely advertising,which will be considered as the
very first example, albeit quitesuperficially at this stage. However,
the reader may already see apossible benefit of the logisticalapproach, such as the generation of
a logistical language, a formulationof goals that would permeate all
divisions of the manufacturing firm,and the provision of qualitativeframeworks. We are among those
who believe that qualitative
analysis is important when it is notdone merely as a gesture. After all,
arent qualitative improvementsoften far superior to quantitative
ones? 2
Logistics everywhere :
analysis and the thesis ina nutshell
Having taken a vaguely reductionist
ap pr o ac h , w hi c h , l o o sel yspeaking, condenses logistics to
the optimization of certain flows,we had to alter our approach whenpassing from pre-SCM to SCM.Considering SCM forced us to refer
to product and information flows
rather than just to product flowswith regard to what needed to beanalyzed and then optimized. This
is where the yes and noanswers to the question of whether
or not SCM has brought a radicalchange to the field of logisticsmanagement converge. If one
believes that the flows referring toproducts in pre-logistics management
are essential, then one has to
consider SCM as a revolution.Otherwise, it becomes more of a
quantitative progress, where oneoptimizes over as many parametersas the available technologies
permit and as far as the efforts ofoptimization provide returns thatbenefit the shareholders.
The point of view we have adopted
and pushed to its limits as the basisof our thesis is that to some extent,the logistics management approach
can be omnipresent in the way onetackles all sort of problems, fromthose found in the sciences to
those encountered in management.Moreover, while restricting ourselves
to the issues covered in this paper,we believe that most, if not all, ofthe issues relating to decision support,
organization, processes and indeedeverything pertaining to the
running of the manufacturing firm,both as an individual entity and aspart of a wider web of networks,
can benefit from the language andmethods of logistics management.In particular, we believe that the
transition to SCM was mostly dueto the emergence of technologies
permitting the processing of biggerand bigger optimization problemsand the collection of more and
more information that could help in
the optimization process. But weintend to go much further and, in
what we think is a natural nextconceptual step, we propose to
tackle most other problems inmanagement as if they wereproblems in logistics management.
Once the other aspects of the life ofa firm, or of networks of firms, have
been dealt with in this way, weexpect that a large scale integration
of the scope of logisticsmanagement will take place, but wewill refrain from making any
attempt to do that here: the bricksmust be made before a wall can bebuilt. In fact, we will limit ourselves
for the time being to providing twoexamples that will hopefully shownthe broad community of people
interested in all aspects ofmanagement an illustration of the
sound basis for our ambitiousprogram. Our aim for now is torepresent certain classical problems
as examples of logistics. A widerange of expertise may be neededto describe how to manage these
logistical problems by themselves,even before the grand unification
across the board takes place. It maybe the case that the logisticalapproach only brings qualitative
changes to certain problems. Webelieve that this may still be quitevaluable, and we will begin with the
field of advertising that mayprovide one of the toughest challenges
to our thesis. We expect that if wecould stimulate even the possibilityof a discussion about advertising as
an example of logistics, it wouldprovide some justification for thevalidity of our proposal.
Before we get to the first example,
we want to state our belief that itwould be unfair to condemn ourapproach just because a disciplineonly permits a qualitative discourse,
and for instance, only qualitativeoptimization after being reinterpreted
as a form of logistics. In fact, evenin the most traditional applications
of logistics management, to reduceits impact to optimization in themost mathematical and formal
senses would be an importantmistake. Thinking about using
empty trucks for new services on
40Supply Chain Forum An International Journal Vol.4 - N1 - 2003 www.supplychain-forum.com
2. This is the idea defended notably by
Wil liam s (1986, 1990), where he indicates that
when confronted with combinatorial problemsthat are too complex, mathematical
optimization fai ls because the computing
power that w ould be needed is way beyond our
reach, and in some cases will be forever so
(the computation would last many tim es the
life of the uni verse with the fastest machi ne
that could ever be buil t). Mathematical
optimization must then give way to some form
of heuristics (which may then possibly use
mathema tics). Willi ams (1986, 1990) il lustrates
this point in the context of the Traveling
Salesman Problem alr eady mentioned. We are
indebted to John Tomli n for telli ng us about the
work of Paul Willi ams after this paper was first
circulated, and to Paul Willi ams for helping us
gain access to hi s work so quickly. Some of
Paul Willi amstheses do indeed seem rathersimil ar to ours, but he promotes a perspective
of optimization everywhere, a point ofview that now has many supporters, rather
than our position of logistics managementeverywhere, which m ight assume a smallereverywherebut seems to not have beenexpli citly advocated before.
-
8/12/2019 Logistics Everywhere
6/11
41Supply Chain Forum An International Journal Vol.4 - N1 - 2003 www.supplychain-forum.com
the return journey, or imagining newshapes of parts to improve packagingby allowing much better weight-to-
volume ratios are just someexamples of how qualitative inputs,
that are facilitated by formulating theright problems, can have dramaticimpacts. Therefore, the strategy
which we propose to use in order toshow the benefits of our thesis innew disciplines is to map the
principles of logistics managementas completely as possible to the new
discipline (or vice-versa), and inparticular, to try to make an earlyidentification of the quantities that
need to be optimized and of theconstraints that are to be found.
Advertising as an exam pleof logistics
In order to represent advertising as a
logistics problem, one needs to thinkof: (1) the message as the equivalentof the product to be transported; (2)
the minds of the people who are thepotential customers as the points of
delivery (e.g., for certain products
for which a recognition is sought
beyond the existing customer base),and (3) the advertising media as the
vehicles of transportation. Of course,it may be appropriate to distinguishbetween the message that was
intended and the message that wasreceived. This is good news for us! Infact, we interpret this example of
discovering a natural distinctionbetween messages as being aninstance of a phenomenon that we
expect to encounter quite often.In the case at hand, thanks to such a
rethinking, the intended messagebecomes the product to be delivered,
while the slogans, trademarks, etc.,become the package. This may haveimplications regarding the relativeimportance of the elements which
constitute the problem and forthe ways they are dealt with. Oneshould remember that in complex
logistical problems: those that arenot examples of simple linear
optimization problems, substantialchanges in approach are sometimesnecessary before one can proceed
with any form of systematicoptimization. For instance, building avery expensive new road may be the
best solution in certain typicallogistical problems. We leave thedescription of any novel advertising
solutions to the experts, but existingpractices such as advertising during
the Super Bowl or the World Seriesmight be considered as theadvertising equivalent of that sort of
expensive road building. We willdevote the next few lines toexamining whether one can identify
constraints and objective functionswhich would permit a more effective
mapping of advertising to thelogistics framework. Essentially, wehave shown so far that the parallels
exist, but to little commercial or evenconceptual advantage. Further
consideration reveals that we may
need metrics to even start viewingadvertising as a case foroptimization.
Arriving at a point where metrics can
be used may be the hardest task eachtime one wants to seriously applylogistics management methods to a
new discipline. If performances ofthe whole system are the onlymetrics that are easy to apply, then
empirical trials are a possible meansto derive benefit from thelogistical/optimization approach. For
instance, the methods found inexperimental economics could be
used, permitting the measurement ofthe performances of logisticsmanagement methods in a new
discipline. Then, the correlation ofthese measurements to acceptedsigns of success could be evaluated.
However, there might be cases, withadvertising being possibly one of
them, where one should not be tooambitious to begin with, and insteadbe content with using a logistical
perspective to help teach andotherwise communicate aboutcertain fields of human thought or
activity, (although Remark 1, below,is much more optimistic). We may
also find that in cases where tasksare accomplished instinctively, butwhere hard decisions also have to be
made at times, the language oflogistics might allow experts andprofessionals to better formulate
their responses to the questionswhich confront them. It also seems
reasonable to expect that unifyingthe language will help communication
within the firm as well as betweenfirms.
From know ledge management
to logistics
Knowledge management (or KM), asit is known, at least in the circles of
information technology (IT), whereeverything of any importance isreferred to by an acronym, is an
interesting field. Facilitating accessto knowledge when it is neededfor the good of the firm or to
benefit humankind, or even moreambitiously, to help generate the
knowledge needed to solve
previously unsolvable problemssounds like an impossible dream.
Who would not want themanagement of knowledge to work
Conjecture: A fundamental
reinterpretation of a problem as
a logistical issue should often
necessitate a rethinking of
its definition in order to map the
discipline in question to a
f ra me wo r k o f l o g i s ti c s
management. Such attention
devoted to the mapping of the
problem would improve the
benefits of the logistical
approach.
Remark 1: As mentioned
previously, we have learned ofwork that uses optimization
techniques to improve the
distribution of advertisements on
the World Wide Web (see, for
instance, Langheinrich et al.
[1999], and Tomlin [2000]). John
Forrest told us that he was involved
in trying to use optimization
techniques to distribute advertising
in the UK as early as 1967.
Magazines were the vehicles used
for the advertisements in question,
but the results were unsatisfactory,
probably because of the limited
power of the computers available
at that time. This work was not
preceded by a full reformulation of
the advertising problem in terms of
logistics management, but it
certainly helps support the idea
that our proposed approach most
probably has a direct commercial
value in addition to its
philosophical implications and to
the way it may facilitatecommunication inside the firm and
between firms.
-
8/12/2019 Logistics Everywhere
7/11
42Supply Chain Forum An International Journal Vol.4 - N1 - 2003 www.supplychain-forum.com
better? KMs place in the toolkit of
the trend setting executive probably
disappeared in 2001, after enjoying a
period of tremendous popularity.
This followed the publication of two
books in 1995 whose titles signaledthe initial hopes for the success of
the new field: The know ledge-creating
company (Nonaka and Takeuchi,
1995), and Wellsprings of k nowl edge
(Leonard, 1995).
A snapshot of how the field evolved
can be seen in the reprinted
collection of some major
management articles on the subject
of KM (Harvard Business Review,
1998), starting with an early article
by Drucker (1988). The author
claimed that within twenty years, thetypical (large) business would
be knowledge-based. Contemporary
with that review, let us also mention
the monograph Working knowledge
(Davenport and Prusak, 1998). Most
contributors to the KM literature
were also the people who benefited
most directly from KM, e.g., through
conferences, consulting, or teaching.
This may explain why, during its few
years of intense success, KM
depended, not only, as was often
stated, on cultural factors, but alsoon the models for national politics
and economics, the attitudes of Wall
Street, on whether long term success
was expected from the firm, on
whether or not the firms that paid
most for KM consulting had a
strategy which looked beyond the
end-of-quarter bottom line, the
morality in corporate board rooms
and more factors of this kind, which
were mostly hidden from view.
At least one other element thatwould have helped cement the place
of KM has been ignored: namely, an
effective and authentic form of
artificial intelligence (AI). As we will
see, this more than anything else
explains KMs fall from grace (while
less ambitious sub-disciplines such
as unstructured data analysis began
to flourish). A logistical approach
leads us to rethink what should be
the fundamental concepts of KM, as
we shall describe in the first
paragraph. In the second paragraph,
in terms of what we consider to befundamental concepts, we shall re-
examine the usual building bricks of
the KM discourse: data, information,
and knowledge. Then, in the third
paragraph the logistics approach is
explicitly invoked, although we
consider that the approach has
already been beneficial in the way
the logistical perspective conditionsthe whole problem (see Remark 5
below, in particular).
Fundame ntal concepts
The basic objectives of KM first need
to be defined. The list can be
extended, but it has to contain the
following three elements, to which
we will restrict our analysis: data,
information, and knowledge. Indeed,
several authors also recognize that
some classification of these three
basic elements is a necessaryprerequisite to any serious
discussion. Our approach is to
consider that these three elements
are not fundamentals, but refer to
some real or also perhaps to some
virtual or imaginary universe. We
also consider that these elements
depend on who attempts to describe,
understand, exploit, improve, or in
general transform such a universe,
and for what reasons. The underlying
rationale for this is that we are
not only interested in the worldo f b u s i ne s s , b u t a l s o i n
interdisciplinary issues (even if we
are not considering the physical and
natural sciences in this paper). The
importance of data, information and
knowledge is not restricted to the
business world and we want to
comprehend their wider significance.
Here are the elements we consider
necessary to define the basic
concepts that need to be governed
by KM:
World elements. These compriseobjects (including living entities
and people), or at least the
objective aspect of them, which in
some cases may be all that is
accessible or necessary. Also
included are entities that allow a
spatial description: (1) reports or
stories that need space-time
descriptions, with a specific time
span, (2) events or records of
events which are essentially
instantaneous, and (3) a mixture of
the above, which might be needed
to describe a personality, forexample.
People and groups. Groups have
concurrent interests, but people in
groups may also have divergent
interests, some supporting thegroups goals and some quiteindependent of them. There may
be neutral entities, and groups orindividual whose position is wellestablished but unknown to some
or all of the other parties. Groupsusually come with explicit
structures that are partly orcompletely hierarchical, althoughsometimes these may be hidden.
Cultural models. These includepolitical trends, opinions, public orsecret agendas, etc., which directly
condition the ways of thinking and
the actions of both the sending and
receiving parties. Messages sent and received . It is the
communication between partiesthat allows them to develop theirinterests in a particular direction.
Messages should also bedistinguished according to theirformal contents, intended contents
and perceived contents. Scales. This refers to the scales,
precision levels, and/or granularityat which one expects tounderstand a problem.
Revisiting data , inform ationand know ledge
At this early and fundamental stage
of the logistical approach to KM, wefeel that from the above list, the
world elements and scales are thecrucial dimensions on which to base
the analysis. This explains why weplace so much emphasis on the pairWeS (World element, Scale). A strong
indication of its importance is thefact that we believe it allowed us toarrive independently and quickly at
concepts central to KM, whilebringing to light ideas which may go
beyond what has been done before.
The data set attached to a WeS is the
collection of everything that isrequired to provide a full
Remark 2: Already, the reader
will probably detect some effects
of logistical thinking in the above
list of basic concepts: a reflection
of the desire to describe the
problem in terms that facilitate its
mapping to the framework of
logistics.
-
8/12/2019 Logistics Everywhere
8/11
43Supply Chain Forum An International Journal Vol.4 - N1 - 2003 www.supplychain-forum.com
description of the world element inquestion at a given scale. A dataelement, data point, unit of data, or,
sometimes-just data, is an atom ofdescription. For instance, a stockname, a time and a price say
$102.23 of the Alpha stock at thattimeforms a data set, in fact a data
set of a special kind, as it fullydefines a WeS. In this example, theprice ($102.23) is a data point.
However, the digit zero is not data,while the collection (0, third digit indecimal notation) is data, of limited
value, which is contained in theprice.
A more ordinary data set, as it isdifferent form the WeS, is attachedto a bottle of milk at a macroscopic
scale. This data set comprises abottles shape, the volume it
contains, the volume it is supposedto contain, its condition andcleanliness, whether or not the
closing device is intact, theproduction date and the expiry
date. To this, one could add anydescription of the product (e.g., is it
regular or low fat?), the quantitiesof nutrients, the nutrients listed onthe bottle, the region of provenance
of the particular batch of milk inquestion, the findings of theveterinary examination of the
cows, an assessment of theveterinarians competence and of
the competence of the person
responsible for evaluating theveterinarian, the historical basis for
trusting the judgment of the
veterinarian, etc., and the list isprobably far from complete!Information has several meaningsthat should be clearly differentiated.
Firstly, it is possible to analyze it asa collection of data intended
(by those who select and/or
communicate it) to provide usually
partial, but sometimes full,knowledge about a WeS. Therein
lies a strong cultural dependence.One can also see it as a collection ofdata that is thought (by those who
receive and/or use it) to providesome, again usually partial, butsometimes full, knowledge of a WeS;
and therein also resides a strongcultural dependence. In addition,
information is presented primarilyin the form of a message, or moregenerally, in the form of a vehicle to
store or communicate knowledgeabout a WeS.Finally, when it becomes knowledge,
information can be split into twotypes, passive knowledge (PK) and
active knowledge (AK), a division
that could be the first of further orconcomitant dichotomies as we will
soon discuss:
The PK of an individual or of agroup is the set of data and
information that can be used bythis individual or group. Some ofit may be inaccessible to anyone,
so long as its existence is knownand provided it can be retrieved
when needed; Implicit PK is the knowledge that
could be generated from a given
data and information set. Thisknowledge may be arrived atvery easily or possibly only after
very detailed analysis;
AK is knowledge that can helpdecide, act on or generate new
knowledge.
The three concepts of data,
information, and knowledge as wehave developed them so far are
considered in a different way to theorthodox KM approach. Remember
that we have taken the statement ofDrucker (1988) to be a lemma,
instead of a definition, which is astrong argument in favor of our
approach. The following dichotomybetween tacit and explicit
knowledge is, on the other hand,pretty much central to the
development of KM in the 1990s.The notions of tacit knowledge
versus explicit knowledge wereintroduced by Polanyi (1983) in the1960s, and used by Nonaka (1991)
and Nonaka and Takeuchi (1994,1995) to formulate a theory of
organizational learning aroundthe theme of the conversion of
knowledge between these twoforms. To us, these notions appear
more obscure than but essentiallyparallel to AK and PK, although this
may not be the case for KM expertswho have worked with them for along time.
Calling logistics to the rescue
We want to demonstrate that thelogistical perspective helps gives
KM some structure, and that thispoint of view gives a clear
indication as to why KM failed (orat least has failed so far). What we
really hope for is to help breathenew life into KM (whether or not its
name remains the same), as webelieve (and everyone agrees) that
the management of knowledge willbe essential in years to come, be itonly because so many firms
accumulate massive quantities ofdata and information, often
accompanied by massive quantitiesof knowledge. It is clear that those
who are the most able to make useof some of these commodities will
have a decisive advantage. So wefirst need first to ask the
fundamental question What iswhat? in order to map standard
logistics to KM. Such a mappingexercise should be a basicrequirement for any new incursion
of logistics into the intellectualworld, and it will often be
Remark 4: Receiving any
information changes the degree
of knowledge about a WeS,
possibly to the extent of
revealing the previously
unsuspected existence of that
WeS, or of any type of that
particular WeS. Information can
also change comprehension,
and further high functions of thebrain, but we will make every
effort to avoid discussing such
matters here.
Remark 5: According to
Davenport and Prusak (1998:2),
the remark that information is
data endowed with relevance
and purpose can be traced at
least as far back as the paper by
Drucker (Harvard Business
Review, 1998) that we have
mentioned previously. We fully
agree with Druckers statement
(1988). However we consider it
to be a lemma, instead of a
definition, as we think it fits in
logically with the definitions of
data and information we have
provided.
Remark 3: No one will
probably ever need, nor ever
be able to determine the data
set of any WeS except for
simple ones such as those
given in the first example.
-
8/12/2019 Logistics Everywhere
9/11
44Supply Chain Forum An International Journal Vol.4 - N1 - 2003 www.supplychain-forum.com
necessary to repeat the exercise,sometimes frequently, when trying tointensify the logistical approach. As
well as understanding the logisticalequivalents in the KM field, it will
also be necessary to do this for eachnew field. Here are some elements ofthe mapping:The goods: the data, the
information, and more generallyany element pertaining to
knowledge (some have expirationdates, others long-shelf lives);
The shelves: the heads ofpersonnel, at various levels in the
organizations hierarchy; Other shelves (for goods with long
shelf-lives): the vehicles for the
transmission of knowledge, i.e.,items that contain it;
The packages: the form in whichdata, information, and knowledge
will be stored, retrieved, and shared
or used in some form or another;
The transport vectors: the variousmeans to transport or transfer
knowledge, from lectures tovideotapes and training, etc.
Next, we shall look at the
optimization problem(s) that mayoccur with the elements that are the
equivalents of traditional logisticalentities according to the mapping.
Then, if not sooner, we must
consider the objectives of the
companies involved, the nature ofthese companies, the economic
system to be used, etc. Somequestions arise simply as a result of
trying to discover the companysobjectives as in any logisticalproblem; for instance: Is the firm only working for the
shareholders; is it a familybusiness built to last a few
generations or something inbetween?
Are there any ambitious goals, or isthere a non-financial/non-economicagenda that also needs to be part
of the objective function design?
Does it operate in an environmentof harsh, unadulterated capitalism,or in a more socially oriented
environment? Is the company only interested in
the end-of-quarter bottom line? Ifnot, what are the firms otherpriorities?
Are the long-term aims establishedonly for the sake of credibility andto defend short-term stock price,
or is there a genuine long-termstrategy?3
Depending on a combination of
choices and constraints, an obviousquantity to be optimized could besomething like the profit, the
efficiency, or the long-term averaged
profit of the company. However, athorough consideration of the large-scale components of the objectives isnot enough to help, or to justify the
existence of KM: the personalobjective functions of people in thefirm have to also be taken into
account. In fact, not everyone maywant all knowledge to circulate
freely, especially if freely also meansfor free (i.e., free from benefits to thesources of knowledge). Let us now
mention a fact that is worthy ofconsideration: everyone is happy to
receive knowledge, if they are notobliged to prove how that knowledgewas acquired. However, being a free
source of knowledge may bedangerous, besides being potentiallypainful and unrewarding. It will be
essential to figure out the rightmetrics, rewards, related incentivesand deterrence systems in order for
things to evolve properly.4 So webelieve that the merely thinking in
terms of logistics managementwould bring considerable
improvements to the design of KMstrategies.
What KM seems to have mostseriously failed to fully acknowledgeis that the movement or transfer of
some forms of knowledge, especiallyof the tacit or active kinds, either
from brain to brain, or from brain tomachine (to brain), may range fromdifficult to impossible given current
capabilities and constraints. It alsomay be too costly (and cost becomes
a more obvious issue when one deals
explicitly with optimization issues).Too much of the KM literature has
very strongly advocated that there ismuch more to knowledge than just
data, information, and passiveand/or explicit knowledge. In doingso, the KM gurus have set impossibly
high expectations by promptingpeople to expect much more from
the field than is realisticallyachievable, given current constraintsand limitations. This issue, which
stands as an obstacle to the successof KM, is of course related to an issuementioned previously, namely, the
lack of an explicit description of theobjectives of the companies
involved.5
We believe that no serious
improvement can be made in themost significant aspects of KM before
major progress is seen in the field of
AI. This statement, in fact, seemsalmost superfluous, as it is
knowledge handling, and thereforesomething similar in nature tointelligence that we are talking about.
However, before AI makes realprogress, advances in KM can be
achieved by lowering expectations,and this is happening right now(often using dilute forms of AI such
as automated text analysis to theirmaximum effect given currentcapabilities. Another example is that
many tools have been developed thatallow the employee to find out useful
information about their own firm andabout global issues (mostly using theInternet). Most of this progress on
the effective management ofknowledge is not made under theumbrella of KM.
The packaging of information has
indeed begun to happen usingtechnologies such as the extendedmarkup language (XML), and we
expect that the combination of XML
and modern search engines,combining basic search,
categorization and prioritizationfunctions and substitutes with
3. One expects affordabi lity to be a necessary
conditio n for the firm to seriously consider long
term i nvestments, but it should not be the sole
factor.
4. This has certainly been understood by some,
although it is, to say the least, often very well
hidden in the KM literature.
5. For instance, the more the fi rm car es only
about its shareholders, the more it considers itsemployees as commodities, and the more
anxi ous it is to extract tacit knowledge from
these employees, then the difficulty in
managing thi s aspect of know ledge becomes
greater.
Remark 6: Some firms (e.g.,
technology companies) create
knowledge as value per se, others
create knowledge mostly as a
means to improve the production
of what the company manufacturesor sells (e.g., know-how, work
methods, controls, workflows,
etc.).
-
8/12/2019 Logistics Everywhere
10/11
semantics, could now permit theefficient mapping of the
management of routine butessential knowledge to logistics.
Examples of such knowledge couldbe the identities of accountholders, how the damages on cars
rented for business are covered,who in the company is qualified tospeak about a particular aspect of
strategy, etc. The basicclassification of knowledge that we
have proposed, and the concept ofWeS in particular, will help in thisprocess. With the passing of time,
AI or related concepts will continueto develop, allowing, for instance,the better handling of unstructured
information.6 These improvements
allow complex mappings fromlogistics management to KM to beperformed. It is also clear that theissues in information handling have
already become so complex thatoptimization of this handling has
begun to help, and in fact is nowvital. The issues we haveexpounded in this discourse have
certainly started being usedimplicitly, as the recent progress inmilitary logistics cannot be
explained in any other way. Thisbrings us back full circle to where it
all started from.
Conclusion
For about twenty years, logisticsmanagement has undergone
important developments, both inmanufacturing and retailing firms.
At the same time, it has benefitedfrom a universally acknowledgedacademic recognition, as
corroborated by the proliferation ofinternational conferences and
specialized journals. Europe has, inpart, caught up with the USA, inparticular through the
development of organizationalinnovations regarding products andinformation flows management.
Today, the leading logisticalstrategies seem to be universalisticin nature and some authors are not
afraid to assert that there isundoubtedly one best way to
develop a sustainable competitiveadvantage. Somewhat provocatively,
this paper is aimed at showing that
logistics management is evenbroader in its application than
most, if not all, observers have
dared to propose so far. It offers away of thinking that, in terms of
logistical language, vision andmethods will allow us to
comprehend and solve managerialproblems that may be quite farremoved from the traditional field
of logistics management.
The position developed here might
have been operationallyunworkable until a few years ago
before the computing revolutionbegan, but it is hard to believe thatit will not bear fruit in a world
where the handling of informationhas evolved and improved to suchan extent. Hopefully, the rigid
treelike organizations will be
rearranged into more reasonableand supple structures or intoframeworks lacking rigidstructures using modern
communication methods to moreeffectively perform the roles ofcommunicating information,
knowledge, commands, andfeelings. To some extent, logistics
may help to redesign the veryorganization of the firm, if onefollows the practical implications of
our thesis. Of course, one does notneed to believe that logistics is
everywhere in order to hope forthis to happen (see, for instanceEvans and Wurster [2000]), but we
do believe that this point of viewwill help those involved by allowingthem to consider from a fresh
perspective the challenges theyface today and will face in the
future.
We cannot end without mentioningthat the Integrated Supply ChainManagement (ISCM) that IBM has
adopted, with earnings of $5 billionin the first year of implementation,goes a long way beyond SCM, evenif it still falls short of the ultra-global vision of logisticseverywhere that we advocatehere. The new view of logistics thatISCM represents is now wellaccepted, assuming, in short, thatlogistics is almost everywhere andin any event, is much moreimportant than ever before. Aftersharing its experience with itsmajor customers, IBM is already
proposing a properly scaled-downversion of ISCM to SMEs under thename of Integrated Supply ChainSolution.
45Supply Chain Forum An International Journal Vol.4 - N1 - 2003 www.supplychain-forum.com
6. Some observers argue that anything thatresembles AI is AI, so that, fo r i nstance, speech
recognition i s AI. Although we have taken the
opposite point of vi ew merely for the sake of
clarity, we wil l not enter into thi s debate, whose
outcome has no impact on our thesis.
-
8/12/2019 Logistics Everywhere
11/11
References
Abad, R. (2002), Le grand march:lapprovisionnement de Paris souslAncien Rgime, Paris: Fayard.
Davenport, T. and Prusak, L. (1998),
Working knowledge: how organizations
manage what they know, Boston (MA):
Harv ard Business School Press.
Evans, P. and Wurster, T. (2000) , Blow n to
bits: how the new economics of
inform ation transforms strategy, Boston
(MA) : Harvard Business School Press.
Gourdin, K. (2001), Global logistics
management: a competitive advantage
for the new mill enium, Oxford: Blackwell
Publishers.
Grey, W., Katircioglu, K., Bagchi, S.,Gallego, G., Adelhelm, M., Seybold, D.,
Stefani s, S. and Shi, D. ( 2003), BeyondROI: mastering the art o f supply chain
value, Supply Chain ManagementReview , Vol. 7, No. 2, pp . 20-27.
Harvard Business Review (1998),
Harvard Business Review on k nowl edge
management, Boston (MA): Harvard
Business School Press.
Heskett, J. (1977) , LogisticsEssential tostrategy,Har vard Business Review, Vol.55, No. 6, pp. 85-96.
Jomi ni, A.-H. ( 2001) , Prcis de lart de laguerre, Paris: Librai rie Perrin [ 1st ed.:
1838] .
Lambert, D., Stock, J. and Ellram, L.
(1998), Fundamentals of logistics
management, Burr Ridge (IL):
IrwinMcGraw-Hill.
Lammi ng, R. (1993), Beyond partnership:
strategies for in novati on and l ean supply,
London: Prentice Hall.
Langheinrich, M., Nakamura, A., Abe, N.,
Kamba, T. and Koseki, Y. (1999),
Unintrusive customization techniquesfor Web advertising,Proceedings of theEight International World Wide Web
Conference (WWW8) , Toronto, availabl e
on http:/ / www8.org.
Leonard, D. (1995), Wellsprings of
knowledge: building and sustaining the
sources of innovation, Boston (MA):
Harv ard Business School Press.
Magee, J. (1968) , Industrial l ogistics:
analysis and management of physical
supply and distribution systems, New
York ( NY): McGraw-Hill .
Mentzer, J., ed. (2000), Supply chain
management, Thousand Oak s (CA): Sage
Publications.
Naccarato , B., Yao, D., Lin , G., Kim , K.,
Koeni g, L., Ettl, M., All an, R., Buckl ey, S.
and Bagchi, S. (2000), Extendedenterprise supply chai n management at
IBM personal systems group and otherdivisions,Interfa ces, Vol. 30, No. 1, pp. 7-25.
Nonaka, I. (1991), The knowledge-creating company, Harvard BusinessReview , Vol. 69, No. 6, pp. 96-104.
Nonaka, I. and Takeuchi, H. (1994), Adynamic theory of organizational
knowledge creation, OrganizationScience, Vol. 5, No. 1, pp . 14-37.
Nonaka, I. and Takeuchi, H. (1995), The
knowledge-creating company: how
japanese com pani es create the dynam ics
of innovati on, Oxford: Oxford University
Press.
Polanyi , M. ( 1983) , The tacit dim ension,
Magnolia (MA): Peter Smith Publishers
[ 1st ed.: 1966] .
Richa rdot, P. (2002) , Les Etats-Uni s,
hyperpuissance militaire, Paris:
Economica & Institut de Stratgi eCompare.
Stevens, G. (1989) , Integrating thesupply chain ,Logistics Today, Vol. 8, No.4, pp. 19-22.
Tomli n, J. (2000), An entropy approachto unin trusive targeted advertising on the
Web, Proceedings of the NinthInternational World Wide Web
Conference (WWW9), Amsterdam,
available on http:/ / www9.org.
Underhill , T. (1996), Strategic all iances:
managing the supply chain , Tulsa (OK) :
PennWell Publishing.
Will iam s, P. ( 1986), Optimization isbest,Inaugural Lecture as Professor ofOperational Research, University of
Southampton, Faculty of MathematicalStudi es, November.
Williams, P. (1990), Optimisation andoperational research, Institute ofMathematics and its Applicat ions
Bull etin , Vol. 26, No. 4, pp . 76-85.
46Supply ChainForum An International Journal Vol 4 - N1 - 2003 www.supplychain-forum.com