adapting a causality model of technical systems to...

LUND UNIVERSITY

PO Box 117221 00 Lund+46 46-222 00 00

Adapting a causality model of technical systems to represent socio-technical systems:A Case study from the BoP

Jagtap, Santosh; Kandachar, Prabhu

2010

Link to publication

Citation for published version (APA):Jagtap, S., & Kandachar, P. (2010). Adapting a causality model of technical systems to represent socio-technicalsystems: A Case study from the BoP. Paper presented at International Symposium on Tools and Methods ofCompetitive Engineering, TMCE 2010, Ancona, Italy.

Total number of authors:2

General rightsUnless other specific re-use rights are stated the following general rights apply:Copyright and moral rights for the publications made accessible in the public portal are retained by the authorsand/or other copyright owners and it is a condition of accessing publications that users recognise and abide by thelegal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private studyor research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal

Read more about Creative commons licenses: https://creativecommons.org/licenses/Take down policyIf you believe that this document breaches copyright please contact us providing details, and we will removeaccess to the work immediately and investigate your claim.

https://portal.research.lu.se/portal/en/publications/adapting-a-causality-model-of-technical-systems-to-represent-sociotechnical-systems-a-case-study-from-the-bop(46de2a1f-f2e7-4e15-9326-051537eef046).html

Proceedings of TMCE 2010 Symposium, April 12-16, 2010, Ancona, Italy, ed. by I. Horvath, F. Mandorli and Z. Rusakc© Organizing Committee of TMCE 2010 Symposium, ISBN 978-90-5155-060-3

ADAPTING A CAUSALITY MODEL OF TECHNICAL SYSTEMS FORSOCIO-TECHNICAL SYSTEMS: A CASE STUDY FROM THE BOP

Santosh JagtapFaculty of Industrial Design Engineering

Delft University of TechnologyThe Netherlands

[email protected]

Prabhu KandacharFaculty of Industrial Design Engineering

Delft University of TechnologyThe Netherlands

[email protected]

ABSTRACT

The base of the world economic pyramid, generallycalled the base of the pyramid (BoP), consists of fourbillion people with average daily income of four dol-lars. Over the past several years, the design and de-velopment of products and services for BoP marketshas been investigated by several authors from differ-ent disciplines. While some authors suggest a sys-tems approach to design and develop products andservices for these markets, a little work has been car-ried out in this area. Specifically, no studies havebeen found in the reviewed literature that systemat-ically explain how businesses bring about systemicchanges in BoP markets. This study methodically ex-plains how businesses bring about systemic changesin these markets through the design and develop-ment process. In order to explain these systemicchanges, the study first describes the developmentof a causality-model for socio-technical systems bymodifying a causality-model of technical systems. Itthen explains the developed causality model for theBoP. Lastly, it tests this model by an in-depth in-vestigation of a case study from the BoP. The find-ings systematically explain the need of: gaining anin-depth understanding of the existing BoP systemto bring about desired actions (e.g. gaining profits,satisfying under-served or un-met needs of BoP cus-tomers, etc.); including different stakeholders fromthe BoP in the design and development process; andappropriately modifying existing BoP system beforedeploying products and services. In addition, thisstudy proposes some applications of the developed

causality-model to support practitioners involved indesigning, developing and assessing interventions(i.e. products and services) for BoP markets.

KEYWORDS

Base of the pyramid (BoP), causality, socio-technicalsystems, interventions

1. INTRODUCTION

The base of the world economic pyramid, generallycalled the base of the pyramid (BoP), consists of fourbillion people with average per day income of fourdollars, and over a billion of these people earn lessthan one dollar a day. Most of these four billionpeople live in rural villages, urban slums, or shan-tytowns. Usually these people have little or no for-mal education. These people are hard to reach viathe conventional means of communication and dis-tribution channels. The quality and quantity of prod-ucts and services available to these people is usuallyinferior [29]. Prahalad and Hart [29] state, “Low-income markets present a prodigious opportunity forthe world’s wealthiest companies – to seek their for-tunes and bring prosperity to the aspiring poor”.

Over the past several years, the design and develop-ment of products and services for the BoP has beeninvestigated by several authors from different disci-plines [21, 23, 26, 30, 35, 36]. Designing and de-veloping products and services for BoP markets re-quires addressing issues in these markets, and theseissues are varied. Jagtap and Kandachar [21] syn-thesized the literature on these issues in the BoP.

1833

Their study found that the issues identified in thestudy conducted by the United Nations DevelopmentProgramme (UNDP) [35] are comprehensive and in-clude those identified in other relevant studies. Theseissues are about: how businesses can gain the in-formation on BoP markets; under-developed regula-tory frameworks; poor physical infrastructure; lackof knowledge and skills of BoP customers; and BoPcustomers’ lack of or poor access to financial ser-vices.

In order to address these diverse issues, an integratedapproach using knowledge from different disciplinessuch as technical sciences, social sciences and man-agement sciences has been proposed by Kandacharand Halme [23]. This approach is in line with the sys-tems approach. While some authors have highlightedthe need and importance of systems approach in theBoP, much less work has been carried out in this areaexcepting the studies of Subrahmanyan and Gomez-Arias [33], Whitney and Kelkar [36], and Nielsenand Samia [27]. However, these studies focus on nar-row areas of the BoP. For example, the work carriedout by Subrahmanyan and Gomez-Arias [33] aimsat developing integrated frameworks to explain con-sumption patterns of the BoP-people. Furthermore,there is to date no research in the BoP that aims at de-veloping a generic framework for explaining causalphenomena of how businesses bring about systemicchanges in the BoP through the design and develop-ment of products and services for these markets.

Systems approach has been investigated in areasother than the BoP. Some of the sectors in these areasare healthcare, economics, agriculture, energy, etc.[2, 7, 8, 13, 16, 25, 34]. However, these studies alsodo not focus on causal phenomena behind systemicchanges. These abovementioned sectors and the BoPare socio-technical systems. In contrary, several au-thors have investigated causal phenomena in the caseof a systems approach to technical systems or arti-facts such as bicycle, aero-engine, telecommunica-tions infrastructure, etc. The literature on the theoryof technical systems is rich and can be used to ex-plain the causality in a socio-technical system. It istherefore important to overcome the weaknesses andlimitations of the reviewed literature on the BoP andon the systems approach regarding socio-technicalsystems.

The overall aim of this research is to develop an un-derstanding of how businesses bring about systemicchanges in BoP markets through the design and de-velopment process. This overall aim is divided into

the following specific aims:• to develop a framework to explain the causality in

a socio-technical system;• to contextualise this framework for the BoP;• to validate or test this framework; and• to develop suggestions, using this framework, to

support practitioners involved in designing, de-veloping and assessing products and services forBoP markets.

In this research, the above-mentioned frameworkturned out to be a model of causality.

This paper is structured as follows. Section 1 pre-sented the background, motivation, and researchaims. Section 2 presents the research methodology.Section 3 explains the developed causality model forsocio-technical systems and for the BoP, and Section4 explains the testing of this model. Some applica-tions of the contextualised causality model to supportpractitioners are proposed in Section 5. Finally, Sec-tion 6 sets out the conclusions.

2. RESEARCH METHODOLOGY

In general, research approaches can be divided intothe following two broader categories: (1) data-driven(i.e. inductive approach); and (2) theory-driven (i.e.deductive approach). In a data-driven approach, the-oretical constructs are derived by analyzing the em-pirical data. In the case of a theory-driven approach,theoretical constructs are developed first, and theseconstructs are then tested using empirical data. Manyresearchers agree that in reality these two approachesdo not exist in their ‘pure’ form [6].

The reviewed literature in the BoP-domain suggeststhat so far data-driven approach has been employedin this area. There is to date no research in theBoP that has used the theory-driven approach. Inthis research, we employ the theory-driven approach.There are three major steps in this research approach:(1) develop a causality model for socio-technical sys-tems by adapting a model of causality found in thereviewed literature on technical systems, and contex-tualise this model for the BoP; and (2) test this con-textualized model by using a case study drawn fromthe BoP.

2.1. Selection of the case study

United Nations Development Programme (UNDP)[35] led an initiative called ‘Growing Inclusive Mar-kets’ (GIM). In this initiative, they analysed 50BoP-cases from ten different sectors such as energy,

1834 Santosh Jagtap, Prabhu Kandachar

Figure 1 Achieving a sub-function using working principles that aredeveloped from physical effects and geometricand material characteristics - adopted from [28].

healthcare, etc. and from different countries. In orderto test the developed theory, we randomly selectedone case from the UNDP-study.

3. DEVELOPMENT OF A CAUSALITYMODEL FOR SOCIO-TECHNICALSYSTEMS

3.1. Causality in a technical system

A given function of a technical system is achievedby a physical process, which is realised byphysicaleffectsand the geometric and material characteristicsof the system [28]. Figure 1 shows two examplesadopted from Pahl and Beitz [28]. The quantitativerelationships between the physical quantities allowthe description of physical effects. Several physicaleffects may need to be combined to achieve a func-tion. On the other hand, a given function might beachieved by one of several different physical effects(e.g. a force can be amplified by the lever effect, thehydraulic effect, the wedge effect, etc.).

Hubka [19] uses the termsprocess structureandfunction structureto explain and solve design prob-lems. The process structure is a transformation be-tween the input and output situations demanded bythe problem. The function structure can be explainedin terms of verbs allowing these processes to happen.Sembugamoorthy and Chandrasekaran [32] explainfunctionas the intended response of a device to exter-nal or internal stimuli.Behaviouris an explanationof how such a response is produced, and is describedby a causal chain of states. A state establishes state of

some objects in the world. Chakrabarti [9] proposesfunctionas “a relation (or relations) between at leasttwo situations, describing the measurable responsesof a device to measurable external stimuli”. He pro-posesform as “one or several structural descriptionsof a solution, and could be described by one or a setof situations”. Bell [5] defines the termstructureas“the physical composition of the device; its compo-nents and connections”;behaviouras “how a deviceworks, what it does in terms of its internal proper-ties”; functionas “a device’s behaviour expressed interms of its purpose”; andpurposeas “the need thatthe device is intended to fulfil”. He states that theabove concepts provide abstract description of theprevious concept; for instance, function is abstrac-tion of behaviour.

Pahl and Beitz [28] defineside effectas “functionallyundesired and unintended effect of a technical systemon a human or on the environment”. Chakrabarti andJohnson [10] discuss the issue for identifying side-effects in conceptual solutions. They defineside ef-fectsas “effects whose outputs affect the intended op-erations of a system”. In order to activate physical ef-fects, the right ‘inputs’ and right ‘contextual parame-ters’ are necessary. The authors have provided an ex-ample – “activation of a piezoelectric effect requiresstressing (input) of certain crystals (context)”. Side-effects can be identified by noticing the inputs andcontexts that are available in the situation in whicha system works. From these inputs and contexts thepossible physical effects that might get activated can

ADAPTING A CAUSALITY MODEL OF TECHNICAL SYSTEMS 1835

be identified. These side effects can be considered assecondary effectsof Hubka’s [19] model of the struc-ture of technical processes.

There are multiple meanings and representations offunction, form, design problems and solutions [9].Chandrasekaran and Josephson [12] state, “thereis also quite a bit of confusion between functionand behavior in the literature”. The SAPPhIREmodel provides a rich causal explanation of a phys-ical phenomenon and attempts “to reach a non-arbitrary degree of detail of behavioural explanation”[11]. The SAPPhIRE (State-Action-Parts-physicalPhenomenon-Inputs-oRgan-physicalEffect) modelexplains the relationships between the followingseven constructs:• parts - “a set of physical components and inter-

faces constituting the system and its environmentof interaction”;

• state - “the attributes and values of attributes thatdefine the properties of a given system at a giveninstant of time during its operation”;

• organ - “the structural context necessary for aphysical effect to be activated”;

• physical effect - “the laws of nature governingchange”;

• input - “the energy, information or material re-quirements for a physical effect to be activated;interpretation of energy / material parameters of achange of state in the context of an organ”;

• physical phenomenon - “a set of potential changesassociated with a given physical effect for a givenorgan and inputs”; and

• action - “an abstract description or high level in-terpretation of a change of state, a changed state,or creation of an input”.

Jagtap et al [22] modified Chakrabarti et al’s [11]SAPPhIRE model by proposing an additional con-struct ‘stimuli’ and two additional relationships ‘em-body’ and ‘affect’. They defined the construct ‘stim-uli’ as follows: “input context necessary for a physi-cal effect to be activated in the presence of the rel-evant organs”. Different aspects of an input (e.g.measure of input’s attribute) and/or relationships be-tween inputs create ‘stimuli’. The developed modelis called the ‘Sym-SAPPhIRE’ model of causality(see Figure 2). The Sym-SAPPhIRE model thus pro-vides a rich causal explanation of an action.

The Sym-SAPPhIRE model can be useful in tack-ling the confusion created by the multiple meaningsand representations of the various concepts such asfunction, behaviour, structure, etc. We believe that

Figure 2 ‘Sym-SAPPhIRE’ model of causality - Jagtapet al [22].

this model has integrated the concepts, namely func-tion, behaviour, and structure. The constructs partsand organs explain the structure of a device, and theconstruct regarding the changes in states describe thebehaviour of a device. Regarding the function ofa device, Chakrabarti et al [11] state, “In our view,function is seen as specific, limited, intended aspectsof the rich causal behaviour of artifacts embedded inand in conjunction with the environment in which itoperates, and could be seen as:• State change• Attained, final state• Inputs• I/O (input/output)transformation• Creation of the context for physical effects to ap-

pear, i.e., organs, etc”.

The SAPPhIRE constructs for a system transmittingpower through a shaft are as follows (see Figure 3):• Parts: shaft forms a revolute pair with bearings;• Organs: one degree-of-freedom of motion be-

tween the shaft and bearings, bearings fixed to arigid support;

• Input: torque applied to the shaft;• Physical Effect: Newtonian laws of motion;• Physical phenomenon: rotation of the shaft;• State: shaft in static state, and shaft in rotating

state;• Action: power is transmitted through the shaft.

Action is interpretation of a change of state, anddepends on the interpreter. This above action (i.e.power is transmitted thorough the shaft) is interpre-tation of the change of state by the authors of thispaper.


Figure 3 Transmission of power through a shaft.

3.2. Causality in a socio-technicalsystem

Technical systems can be represented by using theSym-SAPPhIRE model of causality because in thesesystems we know the organs, stimuli, and the phys-ical effects that can get activated in the presence ofthese stimuli and organs. Furthermore, in the designof these technical systems, the information on thesephysical effects is useful. For example, a databaseof physical effects along with the stimuli and organsrequired to activate these physical effects can assistdesigners in selecting appropriate stimuli and organsto achieve desired functions. In a technical system,in addition to desired physical effects, some unan-ticipated physical affects can also get activated. Bycareful analysis of these systems, these unanticipatedphysical effects can be identified, and thereby unan-ticipated changes of state and actions can be noticed.These unanticipated actions are generally called sideeffects. For example, Jagtap et al [22] analyzed casestudies involving deterioration of components fromaero engines. They used the Sym-SAPPhIRE modelof causality to analyse the data in those case studies.The deterioration mechanisms faced by the compo-nents are called side effects.

Solutions developed by businesses in the BoP in-clude social and technical systems. The technicalsystems can include physical products or infrastruc-ture such electricity network, telecommunications in-frastructure, roads, etc. These physical products canbe represented using the Sym-SAPPhIRE model ofcausality. The social systems do follow the laws ofnature (i.e. physical effects); however, we do notknow these laws for such systems. Therefore, wecan not represent social systems using this model.In addition, we can not represent the relationshipsbetween the social and technical systems using thismodel. The lack of knowledge on the laws of nature

in the case of social systems can be tackled by knowl-edge and judgment of stakeholders involved in de-signing and developing products and services. Theseexperts can use their judgment and experience to es-timate the response of a social system to some inputs.

To represent the causality in a socio-technical sys-tem, we simplify the Sym-SAPPhIRE model ofcausality as shown in Figure 4. We call this sim-plified model as the causality model for a socio-technical system (CMSTS). The system in this modelconsists of relevant elements in it such as people,technical systems (i.e. products), and operating pro-cedures. This construct includes these different ele-ments, relationships between these elements, and itsenvironment of interaction. This is illustrated in thelower part of Figure 4. We have not explicitly in-cluded the constructs ‘organs’ and ‘stimuli’ of theSym-SAPPhIRE model. To simplify, we considerthe construct ‘organ’ to be implicitly included in theconstruct ‘system’, and the construct ‘stimuli’ in theconstruct ‘inputs’. The inputs include material, en-

Figure 4 Causality model for a socio-technical system(CMSTS).


ergy, and information requirements to create somechanges of state for a system. The changes of stateare interpreted as actions, and therefore these actionsare subjective in nature. These changes of state andactions can create or affect inputs and system. Inaddition, the changes of state and action can be in-terpreted as inputs for further changes of state. Thecausality model shown in Figure 4 also includes timedimension.

3.3. Contextualising the CMSTS for BoP

In the case of the BoP, businesses deploy their prod-ucts and services to achieve some desired changesof state and thereby actions. These products andservices act as an intervention. In the case of theBoP, the inputsin Figure 4 can be seen as aninter-ventionand thesystemas aBoP system. This BoPsystem consists of different elements such as BoPcustomers, regulatory frameworks, available infras-tructure (e.g. roads, electricity network, etc.), op-erating procedures, relationships between these ele-ments, and the environment of interaction.

An intervention can be comprised of people, prod-ucts, and processes. Businesses design this interven-tion in order to obtain desired actions such as gainingprofits and satisfying under-served or un-met needsof BoP customers. The intervention and the existingBoP system activate some physical effects which inturn bring about some changes in state, which are in-terpreted as actions. The model in Figure 4 is at abroader-level. However, to bring about the desiredactions, there can be a complex network of causalchains.

In order to achieve desired actions by employing anintervention in an existing system, it is crucial to un-derstand the existing system. In the case of the BoP,an in-depth understanding of different elements inthe BoP, relationships between them, and environ-ment of interaction is crucial to design and developan intervention, which can bring about desired ac-tions. This understanding of the existing BoP sys-tem is input to the process of design and developmentof the intervention. This understanding can includedifferent characteristics of the BoP stakeholders, ex-isting state of the physical infrastructure, regulatoryframeworks in the BoP, and relationships betweenthem. In addition, this understanding can provide in-formation on the issues in the existing system. Inthe case of a BoP system, these issues can be under-developed regulatory frameworks, poor physical in-frastructure, lack of knowledge and skills of BoP cus-

tomers, etc. The CMSTS is thus useful to explain theimportance of gaining an in-depth understanding ofan existing system in order to bring about desired ac-tions by employing an intervention.

The changes of state or actions can create or affectsystem and inputs (see Figure 4). This suggests thatin the case of a BoP system, the changes of stateor actions can affect existing BoP system and inter-vention over time. Therefore, the same interventionmay not result into the desired changes of state oversome time period. This implies the need to period-ically update the understanding of the BoP systemand accordingly redesign the intervention to achievedesired changes of state and actions.

In the case of a socio-technical system, the lackof knowledge on physical effects puts an empha-sis on identifying changes of state using stakehold-ers’ knowledge, experience, and interpretation. Asthe desired actions are interpretations of changes ofstate by different stakeholders, it is important to in-clude these stakeholders in the design process. Dif-ferent stakeholders can have different interpretationsof changes in states – that is – they can interpret dif-ferent actions based on their knowledge and experi-ence. For example, the BoP customers can interpretchanges of state due to an intervention to identifywhether or not the intervention would satisfy theirneeds or if they would accept the changes of statedue to the intervention. Therefore, it is useful to in-clude the stakeholders including BoP customers inthe design of an intervention. This will also help tounderstand the existing BoP system as the stakehold-ers from this system can provide their insights andknowledge regarding that system.

Furthermore, the developed intervention can activatesome unanticipated physical effects and thereby cancreate some unintended changes of state and actions.These unintended actions can be called side effects.In the case of a technical system, a database of phys-ical effects along with required inputs and parts forthe activation of these physical effects can help toidentify side effects. However, in the case of a socio-technical system, a database of physical effects cannot be built because of the lack of knowledge regard-ing physical effects in a social system. The knowl-edge and experience of different stakeholders canhelp to indentify these side effects.

Therefore, their involvement in the design and de-velopment of an intervention is useful in identify-ing side effects. In addition, different stakeholders


can have different interpretations of changes of state;therefore, it is likely to identify different actions fora given change of state. Some of these interpretedactions can be undesired for some stakeholders, andtherefore these actions can be considered as side ef-fects. In essence, different interpretations of changesof state are useful in identifying different desired andundesired (side effects) actions. As shown in Fig-ure 4, a change of state or an action can be interpretedas inputs for a further action. Therefore, differentstakeholders can identify different inputs by differ-ent interpretations of changes of state or actions. Theidentification of different inputs can help to identifyfurther possible changes of state and actions whichcan be desired or undesired.

In the BoP, the process of designing and developingan intervention can affect or modify the existing sys-tem. This design and development process can beseen as ‘inputs’ as shown in Figure 5. We call these‘inputs’ as ‘Design and Development Process Inputs’(DDPI). This DDPI can change the existing BoP sys-tem before implementing the developed intervention.The DDPI can change the regulatory frameworks,physical infrastructure, etc. in the BoP. The DDPIis required when the final intervention can be em-ployed after some changes in the existing system. Asshown in Figure 5, the DDPI and the existing BoPsystem modify the existing system and produce thefinal intervention. The modified system and this in-tervention in turn bring about some desired actionssuch as gaining profits and satisfying needs of BoPcustomers.

4. CASE STUDY

This section illustrates the CMSTS by using a casestudy. This case study presents information on aproject where a company called Amanco designedand developed an irrigation system for low-incomefarmers from Latin America.

The unfair prices and commercial intermediariescaused low productivity in the agricultural output ofsmall farmers from the BoP in Latin America. Acompany, Amanco, developed an integrated irriga-tion system for these farmers who produced lemonson their land. This system aimed at increasing pro-ductivity of the agricultural output and at increasingthe efficiency of water use. In order to implement thedeveloped system, the company collaborated withpartners from the BoP. In addition, the company col-laborated with partners providing micro-credit. Thecompany first developed the system in Guatemala

Figure 5 Modifying the existing BoP system prior toimplementing the final intervention.

and then in Mexico. We illustrate our model (i.e.CMSTS for the BoP) by presenting information onthe system developed in Mexico.

4.1. Existing BoP system

Amanco implemented the irrigation system in a com-munity called La Testaruda in Mexico. The com-pany first gathered information on this community.In order to gather the required information, the com-pany selected a social entrepreneur, Arturo Garcıa,the director of the NGO called Sustainable FarmersNetwork (RASA, Red de Agricultores Sustentables).Ashoka, an international civil society organizationwhich promotes social entrepreneurship worldwidehelped Amnaco to select this social entrepreneurfrom RASA. RASA had experience of over 25 years


in rural projects. In addition, the trust and legitimacyof RASA among the community was useful in gain-ing the required information on the community (i.e.existing BoP system). The farmers in the commu-nity were using outdated irrigation methods due tothe lack of capital, and they lacked the awarenessof the importance of renewing the old lemon plants.The productivity of old lemon plants was poor. Thesefarmers sold lemons in local markets or nearest citieswithout reaching larger wholesalers or supermarketchains.

In addition, the small farmers of the La Testarudacommunity felt neglected as the agriculture was nota priority for the government, and the distribution ofpublic resources lacked a pro-poor approach. Priorto the Amanco’s project, small farmers formed a co-operative, which lobbied and achieved transportationinfrastructure and public lighting services in 2005.These small farmers were unable to afford the totalcost of the new irrigation system, public subsidieswere not reaching them, and they lacked the financialcriteria and standards for getting the required capitalfor investing in the new irrigation system

In addition to the information on existing BoP sys-tem, other inputs were necessary in the design anddevelopment process. These inputs, for example,were experience and knowledge of the Ashoka andAmanco. Ashoka’ experience in the field of socialentrepreneurship helped Amanco. Amanco’s experi-ence and knowledge of designing irrigation systemswas crucial in the design and development process.

4.2. DDPI

The design and development process started in 2005.104 hectares of the La Testaruda community’s landwas selected for the project. In order to tackle thelack of coordination and weak social capital in thiscommunity, social gatherings were arranged amongfarmers and their families for collaborative activities(e.g. preparing roads for excavation).

Amanco carried out the hydraulic design and topo-graphic mapping required for designing technical as-pects of the new irrigation system. The companyprovided training to the technicians from the RASAfor creating distribution channels, and for supportingand supervising the installation of the new irrigationsystem. The installation of the new irrigation sys-tem was the responsibility of small farmers. RASAwas responsible for the promotion of the new irriga-tion system. The promotion was achieved throughmeetings and word-of-mouth strategy among small

farmers. In addition, RASA created cooperatives offarmers in order to facilitate the promotion of the newirrigation system.

The development of financial model was carried outby Amanco and RASA. This model consisted 20%down payment by farmers in three installments, 30%microcredit, and 50% public subsidy. RASA facili-tated access to financing channels and public subsi-dies.

The inputs to the design and development processthus were from the BoP system and the traditionalnon-BoP system. For example, the NGO, RASA, hadexperience in the BoP system, and Amanco had priorexperience of traditional non-BoP markets. Therewere issues or conflicts at the interface between thesepartners from different systems. The lack of un-derstanding of low-income markets required an ex-ploratory and learning approach for Amanco to gainthe market information. In addition, lack of com-petencies to enter low-income markets caused diffi-culties for the company. The company had no priorexperience of working with an NGO. Amanco faceddifficulties in convincing the small farmers about thebenefits of the new irrigation system. Specifically,the company found it hard to convince small farmers,who had good access to water, of the potential bene-fits of the new irrigation system. These small farmershaving good access to water experienced high ineffi-ciencies. The company could convince only ten outof 52 small farmers from the La Testaruda commu-nity to renew their lemon plants. Although, the pro-ductivity of the old lemon plants was poor, the famershad an emotional value for these plants.

4.3. Intervention and the modified BoPsystem

The design and development process modified theexisting system and created the final intervention,which is the new irrigation system. The modifiedBoP system resulted into coordination and increasedsocial capital among small farmers of the La Tes-taruda community. Ten out of 52 farmers renewedtheir lemon plants by 2006. The trained NGO couldsupport and supervise installation of the new irri-gation system, and the financial model was avail-able. The final intervention consisted of the new ir-rigation system. Three types of irrigation systems,namely drip irrigation, portable irrigation, and micro-sprinkling, were designed. The price of the irrigationsystem was in the range 2500 to 3000 US dollars perhectare.


4.4. Actions

The intervention and the modified BoP system cre-ated some changes of state and actions. These actionsare as follows.• Agricultural (i.e. lemon) output increased due to

the new irrigation system and the replacement ofold lemon plants.

• Estimates of the representatives of the IrrigationUnit suggest that the income of the small farmerswould increase by at least three times.

• Labor cost and time required for irrigation re-duced.

• About 60% savings in the consumption of waterwere achieved. In addition, the new irrigation sys-tem helped to mitigate the land erosion.

One of the farmers of the La Testaruda communityinterpreted the changes due to the new irrigation sys-tem as, “there is a renovated hope as we are startingto see the transformation”.

5. IMPLICATIONS: SUPPORTING THEPRACTITIONERS

Sections 3 explained the CMSTS, and Section 4 ex-plained its testing by using a case study from the BoP.This section presents some applications of the CM-STS to support practitioners involved in designing,developing and assessing interventions for BoP mar-kets.

5.1. Use of the past and existingBoP-interventions

As mentioned in Section 2.1, United Nations De-velopment Programme [35] led an initiative called‘Growing Inclusive Markets’. In this initiative, theyanalysed 50 BoP-cases from ten different sectorssuch as energy, healthcare, etc. and from differentcountries. Some part of this UNDP-study is reportedby Gradl et al. [15]. For each of the 50 cases, theUNDP-study identified relevant issues considered bythe business and the strategies used to address theseissues. While this UNDP-study has identified issuesin BoP markets and strategies used by businesses toaddress these issues, it has not developed a frame-work to structure the information on the 50 cases.The information on these 50 cases is stored a textformat, and these cases are not represented using acausality framework.

The CMSTS is useful to structure the informationon the past and existing BoP-cases. This can beachieved by using the constructs of the CMSTS (e.g.

existing BoP system, DDPI, modified BoP system,final intervention, actions, etc.). The case study onAmanco’s irrigation system in Mexico, described inSection 4, has illustrated how the constructs of theCMSTS can be used to structure the information ona BoP-case. Furthermore, the CMSTS allows a di-agrammatic representation of information. The re-viewed literature [1, 17, 24, 31] suggests that the di-agrammatic representation of information during theearly phases of a design process has the followingadvantages:• it helps designers to comprehend problems;• it improves the effectiveness and efficiency of

early design, and facilitates human cognitive pro-cesses fostering innovation;

• information can be analysed at a faster rate thanin a text format.

This implies that diagrammatic representation of in-formation on the past and existing BoP-cases canhelp practitioners in designing a new intervention orin redesigning an existing intervention. The CM-STS helps to diagrammatically represent informa-tion. Figure 6 exemplifies the diagrammatic repre-sentation of the information on the Amanco’s inter-vention regarding the irrigation system. This dia-grammatic representation can help practitioners whoare involved in: (1) designing a new irrigation sys-tem in BoP markets; or (2) redesigning an existingirrigation system. It is thus clear that the CMSTS isuseful in structuring and representing information onBoP-cases, and a database of the past and existingBoP-cases can be developed using the CMSTS. Sucha database can help practitioners to: (1) easily com-prehend these BoP-cases; and (2) generate solutionsfor the problems they are tackling.

Information on interventions or cases similar to theone being designed can help practitioners to under-stand different issues relevant to their design task.Analogical thinking/reasoning can assist in identi-fying the similar interventions or cases. Ball etal [3] state “Analogical reasoning entails the useof ‘source’ information from a previous problem-solving episode as a means to facilitate attempts atsolving a current, ‘target’ problem.” Holyoak andThagard [18] explain that “. . . analogical thinking in-volves establishing a mapping, or systematic set ofcorrespondences, between the elements of the sourceand the target analog.” The constructs of the CM-STS (e.g. actions, system, intervention, etc.) can beused to search for similar cases or interventions inthe abovementioned database.


Figure 6 Diagrammatic representation of information using the CMSTS.

A single construct or a combination of constructs(e.g. actions plus intervention) can be used to searchfor similar cases or interventions. For example, in thedesign of a new irrigation system for a BoP market,practitioners might search a database of BoP-casesusing the following criteria:• ‘action’ to increase agricultural productivity (sim-

ple search);• ‘action’ to increase agricultural productivity plus

‘intervention’ consisting of a drip irrigation sys-tem (combination search).

The search results can help the practitioners to gener-ate appropriate solutions to tackle the problems theyare solving.

5.2. Impact assessment

The assessment of the impact of the interventionsaimed at social development is important [4, 14, 20].

The impact assessment (IA) in BoP markets helpsto understand the effects of an intervention on BoP-clients. The information, gained through IA, on theperformance of an intervention is useful in the de-sign and development of a new intervention or re-designing an existing intervention. Furthermore, thisinformation can be useful in commercial promotion,marketing, or to expand the intervention to other lo-cations in BoP markets.

The CMSTS can be useful in IA. This can beachieved by using the following constructs or rela-tionships between the constructs of the CMSTS:• the construct ‘changes of state’;• the relationship ‘interpreted as’ used to arrive at

‘actions’ from ‘changes of state’.

An intervention in a BoP-system brings about somechanges of state. These changes of state reflect the


impact of the intervention. This requires informationon states of a target BoP-system before and after em-ploying DDPI and final intervention. As shown inFigure 7, the initial state (i.e. State 1-target) and fi-nal state (i.e. State 2-target) of a target system canbe compared to assess the impact of an interven-tion. The information on a BoP-system (i.e. con-trol system) where the DDPI and intervention are notimplemented can also help in IA. The control sys-tem needs to be as similar as possible with the tar-get BoP-system. The comparison of ‘State 2-target’with ‘State 2-control’ is useful in assessing the im-pact of an intervention (see Figure 7). It is impor-tant that ‘State 2-control’ is obtained over the sametime period as ‘State 2-target’ is achieved. Differentobjective parameters such as income of BoP-people,number of products and services available to them,etc. can be used to specify a state of a BoP-system.

Figure 7 Using the construct changes of state of theCMSTS to assess impact.

The relationship ‘interpreted as’ of the CMSTS usedto arrive at ‘actions’ from ‘changes of state’ can alsobe useful in IA. This is explained as follows. Some-times, information on ‘State 1-target’ and a controlsystem is not available. In addition, there can be in-sufficient number of objective parameters required tocompletely specify a state of a BoP-system. Undersuch circumstances, the perceptions of the people af-fected by an intervention are useful in IA. These per-ceptions are their interpretations of the changes ofstate due to an intervention. Regarding these per-ceptions in IA, Eaton at al [14] state, “. . . these per-ceptions are what will determine their behaviour inthe future and possibly play as important a role than“actual” performance (of an intervention)”. As men-tioned in Section 4, the positive impact of the inter-vention in the La Testaruda community is reflected inthe perception of one of the farmers of this commu-nity as, “there is a renovated hope as we are starting

to see the transformation”. Different methods suchas surveys, interviews, participatory learning and ac-tion, etc. can be used to identify BoP-clients’ percep-tions or interpretations of the changes of state due toan intervention.

6. CONCLUSIONS

Designing products and services for BoP markets re-quires addressing different issues such as access to fi-nancial services, gaining market information, under-developed regulatory frameworks, etc. In order toaddress these issues, businesses bring about systemicchanges in these markets through the design and de-velopment process. Our study systematically ex-plained how businesses bring about these systemicchanges in BoP markets.

This study simplified the Sym-SAPPhIRE model ofcausality, originally developed for technical systems,to explain causality in socio-technical systems. Inthis simplified model, called CMSTS, the ‘system’consists of relevant elements in it such people, tech-nical systems (e.g. products), and operating proce-dures. We have implicitly included the construct ‘or-gan’ of the Sym-SAPPhIRE model in the construct‘system’ of the CMSTS, and the construct ‘stim-uli’ in the construct ‘inputs’ of the CMSTS. We alsoadded time dimension in the CMSTS.

In addition, this study contextualized the CMSTS forthe BoP. In the case of the BoP, the inputs to the BoPsystem can be seen as an intervention designed anddeveloped by businesses for these markets. This in-tervention can be comprised of people, technical sys-tems, and processes. The intervention and the exist-ing BoP system activate some physical effects, andthereby bring about some changes in state, which areinterpreted as actions (e.g. increased profits, satisfy-ing un-met or under-served needs of BoP customers).

In the case of BoP markets, the process of designingand developing an intervention can affect or modifythe existing system, and this process can be seen as‘inputs’ called DDPI. This DDPI changes the exist-ing BoP system before implementing the developedintervention.

This study tested the CMSTS for the BoP by usinga case study, namely, Amanco’s integrated irrigationsystem in Mexico. However, there are limitations tothis study. We tested our CMSTS for the BoP usinga single case study. A larger number of cases canhelp refine this model to increase our understandingof how businesses bring about systemic changes in


the BoP by designing and developing products andservices for these markets.

Our study has important practical implications. Thisstudy systematically explains the importance of:• gaining an in-depth understanding of the exist-

ing BoP system to bring about desired changesof state and actions;

• including different stakeholders from the BoP inthe design and development process;

• appropriately modifying existing BoP system be-fore implementing the final intervention.

Furthermore, the CMSTS can be used to support thepractitioners involved in designing, developing andassessing interventions for BoP markets. The CM-STS can be useful in structuring information on thepast and existing BoP-cases, and this information canbe used in the design and development of a new in-tervention or in the redesign of an existing interven-tion. In addition, the CMSTS’s construct ‘changesof state’ and the relationship ‘interpreted as’ used toarrive at ‘actions’ from ‘changes of state’ can help inassessing the impact of an intervention.

This study explored the causality in BoP markets ata system-wide level. While studies at micro level fo-cusing on some particular issues in the BoP are im-portant, they may provide insufficient attention to theinteractions between different parts of a broader sys-tem. We believe greater effort should be made tocarry out research in the field of BoP markets to un-derstand different aspects at a macro or system-widelevel.

ACKNOWLEDGMENT

We would like to thank the Delft University of Tech-nology, Netherlands for funding this research. Wewould also like to thank Christina Gradl for provid-ing the data on the case study.

REFERENCES

[1] Aurisicchio, M., Bracewell, R.H., and Wallace,K.M., (2007), “Characterising Design Ques-tions That Involve Reasoning”, Proceedings ofthe International Conference on EngineeringDesign, ICED’07, Paris, France.

[2] Azapagic, A., (2003), “Systems Approach toCorporate Sustainability: A General Manage-ment Framework”, Process Safety and Envi-ronmental Protection, Vol. 81 (5), pp. 303-316.

[3] Ball, L.J., Ormerod, T.C., and Morley, N.J.,(2004), “Spontaneous Analogising in Engi-

neering Design: A Comparative Analysis ofExperts and Novices”, Design Studies, Vol. 25(5), pp. 495-508.

[4] Becker, H.A., (2001), “Social impact assess-ment”, European Journal of Operational Re-search, Vol. 128 (2), pp. 311-321.

[5] Bell, J., (2004), “Representation of Function”,University of Wales, Aberystwyth, UK, 2004.

[6] Blessing, L.T.M., and Chakrabarti, A., (2009),“DRM, a Design Research Methodology”,Springer-Verlag London Limited, London

[7] Boston, J., (2000), “The challenge of evaluat-ing systemic change: the case of public man-agement reform”, International Public Man-agement Journal, Vol. 3 (1), pp. 23-46.

[8] Buchholz, T.S., Volk, T.A., and Luzadis, V.A.,(2007), “A participatory systems approach tomodeling social, economic, and ecologicalcomponents of bioenergy”, Energy Policy, Vol.35 (12), pp. 6084-6094.

[9] Chakrabarti, A., (1993), “Towards A Theoryfor Functional Reasoning in Design”, Proceed-ings of the International Conference on Engi-neering Design, ICED’93, The Hague.

[10] Chakrabarti, A., and Johnson, A., (1999), “De-tecting Side Effects in Solution Principles”,Proceedings of the International Conference onEngineering Design, ICED’99, Munich.

[11] Chakrabarti, A., Sarkar, P., Leelavathamma, B.,and Nataraju, B.S., (2005), “A Functional Rep-resentation for Aiding Biomimetic and Artifi-cial Inspiration of New Ideas”, Artificial Intel-ligence in Engineering Design and Manufac-turing, AI EDAM, Vol. 19 (2), pp. 113-132.

[12] Chandrasekaran, B., and Josephson, J.R.,(1997), “Representing Function as Effect”,Proceedings of the Proceedings of FunctionalModeling Workshop, Electricite de France,Paris, France.

[13] Duczynski, G., (2004), “Systems approachesto economic development for indigenous peo-ple: a case study of the Noongar Aboriginals ofAustralia”, Futures, Vol. 36 (8), pp. 869-888.

[14] Eaton, D., Wiersinga, R.C., and Danse, M.,(2009), “Impact of high quality vegetable seedsfor smallholder farmers in South East Asia”,Proceedings of the International Conference onImpact of Base of the Pyramid Ventures, Kan-dachar, P., Jongh, I.d., Halme, M. (Eds.), Delft,Netherlands.


[15] Gradl, C., Sobhani, S., Bootsman, A., and Gas-nier, A., (2008), “Understanding the markets ofthe poor: A market system approach to inclu-sive business models”, in: Sustainability Chal-lenges and Solutions at the Base of the Pyra-mid, Kandachar, P., Halme, M. (Eds.), Green-leaf Publishing Limited, Sheffield, UK.

[16] Harrison, M.I., Koppel, R., and Bar-Lev, S.,(2007), “Unintended Consequences of Infor-mation Technologies in Health Care—An In-teractive Sociotechnical Analysis”, J Am MedInform Assoc., Vol. 14 (5), pp. 542–549.

[17] Hoffman, R.R., Coffey, J.W., Haye, P.J., Canas,A.J., Ford, K.M., and Carnot, M.J., (2002),“One Small Step for a Diagram, One GiantLeap for Meaning”, Proceedings of the Dia-grammatic Representation and Inference : Sec-ond International Conference, Diagrams 2002,Callaway Gardens, GA, USA.

[18] Holyoak, K.J., and Thagard, P., (1995), “Men-tal Leaps: Analogy in Creative Thought”, MITPress, Cambridge, MA, USA.

[19] Hubka, V., (1982), “Principles of EngineeringDesign”, Butterworth Scientific.

[20] Hulme, D., (2000), “Impact AssessmentMethodologies for Microfinance: Theory, Ex-perience and Better Practice”, World Develop-ment, Vol. 28 (1), pp. 79-98.

[21] Jagtap, S., and Kandachar, P., (2009), “To-wards Linking Disruptive Innovations and BoPMarkets”, Proceedings of the InternationalConference on Engineering Design, ICED-09,Stanford, California.

[22] Jagtap, S.N., (2008), “Capture and Structureof In-Service Information for Engineering De-signers”, Thesis (PhD), University of Cam-bridge, U.K.

[23] Kandachar, P., and Halme, M., (2008),“Farewell to pyramids: how can businessand technology help to eradicate poverty?”,in: Sustainability Challenges and Solutionsat the Base of the Pyramid, Kandachar, P.,Halme, M. (Eds.), Greenleaf Publishing Lim-ited, Sheffield, UK.

[24] Kokotovich, V., (2008), “Problem Analysis andThinking Tools: An Empirical Study of Non-Hierarchical Mind Mapping”, Design Studies,Vol. 29 (1), pp. 49-69.

[25] Leveson, N., Dulac, N., Marais, K., and Car-roll, J., (2009), “Moving Beyond Normal Ac-cidents and High Reliability Organizations: A

Systems Approach to Safety in Complex Sys-tems”, Organization Studies, Vol. 30 (2-3), pp.227-249.

[26] London, T., Anupindi, R., and Sheth, S., (2009)“Creating mutual value: Lessons learned fromventures serving base of the pyramid produc-ers”, Journal of Business Research, Vol. InPress.

[27] Nielsen, C., and Samia, P.M., (2008), “Under-standing key factors in social enterprise devel-opment of the BOP: a systems approach ap-plied to case studies in the Philippines”, Jour-nal of Consumer Marketing, Vol. 25 (7), pp.446 - 454.

[28] Pahl, G., and Beitz, W., (1996), “EngineeringDesign”, 2 ed., Springer-Verlag, London.

[29] Prahalad, C.K., and Hart, S.L., (2002), “TheFortune at the Bottom of the Pyramid”, strat-egy+business, 2002.

[30] Prahalad, C.K., (2004), “The Fortune at theBottom of the Pyramid: Eradicating Povertythrough Profits”, Wharton School Publishing,Upper Saddle River: Nj.

[31] Salustri, F.A., Weerasinghe, J.S., Bracewell,R.H., and Eng, N.L., (2007), “Visualisingearly engineering design information with di-agrams”, Journal of Design Research, Vol. 6(1-2), pp. 190–217.

[32] Sembugamoorthy, and Chandrasekaran, B.,(1986), “Functional Representation of De-vices and Compilation of Diagnostic Problem-Solving Systems”, in: Experience, Memoryand Reasoning, Lawrence Erlbaum Associates,Hillsdale, NJ.

[33] Subrahmanyan, S., and Gomez-Arias, J.T.,(2008 ), “Integrated approach to understand-ing consumer behavior at bottom of pyramid”,Journal of Consumer Marketing, Vol. 25 (7),pp. 402 - 412.

[34] Temel, T., (2005), “A systems approach tomalaria control: an institutional perspective”,Health Policy, Vol. 71 (2), pp. 161-180.

[35] UNDP, (2008), “Creating Value for All: Strate-gies for Doing Business with the Poor”, UnitedNations Development Programme, 2008.

[36] Whitney, P., and Kelkar, A., (2004), “Design-ing for the Base of the Pyramid”, Design Man-agement Review, Vol. 15 (4).


1846 This page is deliberately left empty.

adapting a causality model of technical systems to...

Documents