2013 IEEE International Conference on Systems, Man, and Cybernetics
Multi-Modeling : A Systemic approach to address complex scenarios Case illustration for business conflict resolution
Ravi Shankar P
Supriya Kummamuru
Business Systems and Cybernetics Center Tata Consultancy Services (TCS) Hyderabad, India
Business Systems and Cybernetics Center Tata Consultancy Services Hyderabad, India
[email protected]
[email protected]
Abstract - The paper presents a case application of the Multi modeling approach to solve a complex tribal welfare problem. The case in context was a Tribal Welfare organization trying to meet conflicting objectives relating to welfare and Profit. Complex problems comprise multiple aspects to be understood and addressed as a whole. We often deal with organizational complexity by presuming that it does not simply exist. Multimodeling stands against this general trend, in being based on the belief that organizations in the current environments are generally complex, ambiguous and paradoxical. Multi-modeling acknowledges that the real challenge is to learn to deal with this complexity. Multi-modeling offers a way in which this challenge can be taken on by allowing us to grasp and deal with the manysided character of organizational life and in doing so, the multimodeler finds new ways of organizing thoughts and creative ways of approaching and solving organizational problems. The paper introduces the multi modeling approach and illustrates this in a comprehensive manner through a case study.
make the solution holistic since it addresses the multi dimensionality inherent in problems. Here we understand multi dimensionality as the presence of multiple aspects of the systems that are affected by the problem and hence look at how different models can be used to understand these different aspects of the problem. It looks at the aspects of the system as a whole like property of viability, living, hierarchy, emergence which different parts of the system put together display as opposed to the parts in isolation. The approach is based on the premise that systems act as a whole and we need to understand them also holistically. The other interpretation of Gharajedaghi is that ‘Multidimensionality [5] probably is one of the most potent principles of systems thinking. It is the ability to see complementary relations in opposing tendencies and to create feasible wholes with infeasible parts’ is also considered. Also Gökçe Sargut and Rita Gunther in their article (3) propose three properties to determine the complexity of an environment first multiplicity, refers to the number of potentially interacting elements, second interdependence, relates to how connected they are and third, diversity, which deals with the degree of their heterogeneity. Greater the multiplicity, interdependence, and diversity, greater is the complexity associated with the system. The relevance of this approach, of finding solutions to the various aspects of the problem or multidimensionality, is done by the authors and by other consultants in various domains. The introductory Section is followed by the Philosophy of multi modeling in Section II, Overview of the problem solving process encompassing the approach in Section III followed by the case study in Section IV ending with conclusions.
Keywords-Multi Modeling; Complexity; Organizational problems; conflicting objectives, solving problems, case study (key words)
I.
INTRODUCTION
Complex problems constitute one of the major areas of concern in the contemporary world. The behavior of complex systems is often counter intuitive, defying the linear causal interpretation normally employed in problem solving. To handle these systems, there is need for new "tools of thought". Modelling proves very beneficial in discovering such tools and learning how to use them [1] It is necessary to have a method of inquiring into a problem to know the truth of the system under study. Since ages this has, been a subject of study as it has two components: the ontology and epistemology, the nature of the system and how to acquire the knowledge, respectively. In our case the ontology is that the system or Businesses are complex [2, 3]. The approach is based on the premise that complex systems are multidimensional in nature and are best understood by the use and application of multiple models for addressing various perspectives or aspects of the problem. This is broadly known as Multi modeling [4] and has been usefully deployed in many instances to solve problems in differing applications. Multi modeling presents an excellent way to 978-1-4799-0652-9/13 $31.00 © 2013 IEEE DOI 10.1109/SMC.2013.60
II.
MULTI MODELLING
A. The Philosophy Most of us have read the fable of the six blind men of Hindustan. Each felt one part of the Elephant and each one had their own perception about what it was. Long and bitterly they argued, each sticking to his own view, until, unable to resolve the issue, they applied to the elephant keeper for judgment. The
316
processes multiple systemic models are used. The models used help answer the questions of each process. The same model may run across all the processes answering different questions.
wise man led each on a touch-tour of the elephant, until each felt what all the others had felt and to each he asked, "Do you feel what the others felt?" and each replied "yes, I do!""Know then, said the wise man that the nature of this elephant is rope and snake, wall and fan, horn and tree and yet it is more than just all these!" And the six blind men went on their way, chastened and a little less blind than before.
In multi-modeling, the problem context is viewed as a 'mess' rather than a well-defined problem. This diagnosis accepts that the consultant consciously expands his/her focus or diverges to include apparently conflicting contradictory perceptions regarding the situation and establish the linkages between them. There is then a need to systematically converge to identify the purpose of the client system and the associated critical issues of concern. This is the Discovery phase of the consultancy process.
This fable is particularly relevant to the consultancy situation which essentially tries to re-construct an organizational situation from different sources of information. If the situation is compared to the elephant in our fable, we can make some interesting observations here. We also try to manage complexity by focusing on only one small area at a time. The performance of a complex system depends on how all the parts work together, not how each part performs when taken separately. In order to do this we need to employ multiple lenses which will help us view the problems in totality. If the lens is clear to us we can view the model through them. Past research has shown that the multiple characteristics of a whole system can be understood by different models applied on it. For instance, the use of the Viable Systems Model developed by Stafford Beer helps us to diagnose the systems viability. [6]. Miller’s seminal work on Living Systems Theory [7] is a general theory about how all living systems "work", about how they maintain themselves and how they develop and change. The model proposed by him can be used to understand the living properties [7]. The adoption of the multiple models through the process of systemic inquiry has been developed in the Multi modeling approach. The use of multiple models helped in illuminating various facets of complex problems. Cybernetic Influence Diagram (CID™), Stakeholder Analysis Model SNAC™ etc and equivalent control systems models, provide clues to the diagnosis of the problem [8, 9]
The next phase is Diagnosis. The client system has to be analyzed to identify the underlying causes and the barriers that prevent it from functioning effectively. This has to be done with reference to the purpose of the client system as defined and the problem areas identified. The Design phase involves the generation and design of interventions that have to be undertaken to overcome the above barriers. Subsequently, a clear plan for implementation is drawn up and mechanisms for monitoring the same are designed. If the intervention throws up new barriers, the problem areas may need to be redefined. Thus consultancy essential is a progressive refinement and restructuring of the problem context through a process of continuous iteration of discovery, diagnosis and design phases. Multi-modeling seeks to look at client systems on the basis of as many perspectives (or images) as possible. These perspectives translate into formal models in many cases. Something like Stakeholder Model (SNAC) [10], which basically views the client systems as a cultural system, is as valid and relevant as any formal model and may be used in conjunction with any such formal models for maximum understanding.
Thus the problem can be understood and addressed by the use of these different techniques or models suitably to make the solution as holistic as possible. This is the ideal approach to designing sustainable solutions and has often been referred to as ‘Multi modeling’ method of problem solving [8]
There are two loops of Multi-Modeling [11], depicted in Figure1. Logic Loop: Discovery - Diagnosis - Design (Clockwise) Learning Loop: Design - Diagnosis - Discovery (Anti clockwise)
B. The Problem Solving process [10] This section dwells briefly upon the key problem solving processes encompassing the multi-modeling approach. There are three key processes that can be characterized by the answers to the following three questions
The essential flow of the logic loop is in identifying the identity and purpose of the client system, the problem boundary, system diagnosis (barriers to be overcome) and design of interventions (actions to overcome the barriers).
Discovery: What is the nature and identity of the client system and hence, what are the objectives/purposes being pursued by it? Diagnosis: What barriers have to be overcome to achieve these objectives/purposes? Design (of interventions): What actions have to be undertaken to overcome the hurdles/barriers? The processes – termed as Discovery, Diagnosis and Design (of interventions) – that need to be carried out are coterminus, i.e., the crystallization of the three occurs together. This essentially means that while each of the process has its own characteristic nature and activities and outcomes, the processes themselves are iterative in nature. In each of these
Figure 1.0 Logic & Learning Loop
Tata Consultancy Services (sponsors)
317
XC finds the business of buying Essential commodities from the open market at prevalent prices and selling at relative low affordable prices to the poor tribal’s an unviable activity. The reasons for this is that XC has got policies imposed by the governmental regulations that it has to buy through a tendering process only and it has to buy at the lowest prices possible. It is also constrained by the inability to buy larger quantities of the various essential commodities (EC). Storage is also a problem. The only way to make the operation viable is to decrease the cost of purchase and thereby provide adequate margin for the rest of the operations. The problem is to develop a way by which this operation is made viable. In the overall context described above, we discuss a small part of the problem to make the presentation and the understanding of the method lucid. As stated above, the methodology is a sequence of activities and model applications that logically and holistically leads to the solution to the problem. The case was solved using the approach to ‘Problem solving using Multi modeling’ shown in Figure 2.0.The smaller portion of XCs problem that we have worked on to understand and analyze and finally define the solution was as below. Here we intend to define the strategy that will solve the problem of poor margins for XC.
The flow of the learning loop is the other way around: examining whether the interventions throw up new barriers and whether these lead to the need for redefining the purpose / identity and objectives.
Figure 2.0 Steps in Multi Modeling for problem solving It is the interacting and iterative nature of the loops in multimodeling that gives it the variety to match the complexity of the organizational problems. Thus, in principle, at any stage during the consultancy assignment, it should be possible to document the understanding at that stage in the form of three areas which could service as 'internal deliverables' for the consulting team. Figure 2 captures the detailed steps for problem solving adopting the multi modeling approach, outlining the specific steps and associated models which can be used for analysis. III.
B. Problem Statement The Open markets for ECs are sensitive to payment terms. XC’s purchase policies are to pay the lowest prices through the tendering process. XC pays after availing a credit period of 4560 days. The payment process also involves time and effort on the part of the vendor. Hence, prices quoted to XC by vendors are generally higher than to a customer who pays immediately on supply. On the other hand, XC cannot sell at high prices since the Tribal’s buying power is low. Hence margins are dependent on the purchase process effectiveness in terms of getting low prices and needing less storage and transport. Unless margins are good, the viability of the operation (2nd Supply chain) is in question.
CASE STUDY
A. Background This case in study is a tribal welfare organization (XC). Tribal’s living in forests have monopoly rights over all resources grown naturally or otherwise in these identified forests. These resources are huge and valuable consumer items after a little processing XC is a government organization charged with managing a conflicting supply chain for their welfare. It is responsible for performing two key activities: 1. Buy forest produce from the tribal’s at remunerative prices, process and sell them at competitive market prices 2. Buy essential commodities for the tribal’s at wholesale prices and supply to them at almost cost price. XC needs to be maintaining viability while performing these two activities. XC as an organization has over the years become unviable and hence there was a need for a strategy to make it viable and self sustaining.
This problem statement was arrived at by putting together the problem descriptions given by the multiple stakeholders and by using ‘Stakeholder Engagement techniques consisting of one to one meetings and group techniques like ‘the Nominal Group Technique’ [12]. These techniques are used to remove the inconsistencies in stakeholder perceptions and to remove any disagreements. These techniques also help capture the complexities and the specific problem areas that cannot be captured through traditional processes used in Needs gathering. This Problem statement is treated as a reference document for the Problem Solving exercise and is equal to the Voice of the customer.
Traditional understanding of this problem points to nonviability due to high procurement prices, and high operating costs. XC had the task of dealing with these conflicting objectives of making profit on one hand and ensuring welfare for the tribal’s on the other hand. However, welfare is a prime need for the tribal’s and hence the task was to arrive at a strategy to make this a viable, self sustaining Organization.
C. Discover & Diagnose The next step in the method is to understand the problem statement and derive the knowledge that will help us to define an effective problem solution. There could be many ways to understand the problem statement, but the most effective way
318
would be to identify and analyze the interrelationships and the not so visible facts about it. In the Multi modeling approach, we use the Cybernetic Influence Diagram (CID™) for this. In this instance we have used the CID™ which is drawn out of the problem statement. The CID in this case is shown in Figure 3 below, covering the problem statement and forms a subset of the CID for the larger problem of XC.
3.
Similarly Purchase Quantity and Cash available for operations are critical for viability
4.
A preferred condition for viability is the independence of Price and Quantity of Purchase
Understanding the relationships and the Implications of these relationships is an easy work and can be done by identifying the loops and the concentration of inflowing and out flowing arrows on different parameters in the CID™. Some typical verbalizations and their implications are given in the Table 1 below detail our understanding of the Problem. TABLE 1 Loop
6-10-8-5-6
Implications
6-10-9-3-6
Implications
6-10-8-1- 9-85-3-6
Implication
Verbalization
Figure 3.0. CID of XC’s purchase of essential commodities
Margins critical to the operation influence the Cash available for Purchase of ECs. This determines the Credit Period that XC can support and hence the EC Purchase Price that the market will offer. Consequently the Margin resulting for XC is determined
D. Design Every System was intended to meet specific and generic needs, which are specific to stakeholders. In the design of the system, these have been addressed. However, there are constraints that were present in the problem/needs meeting, which have been met when meeting the needs.
The Credit period determines the Purchase Price. Accordingly the Margins from the operation so critical to the Viability of the EC Operation is determined
Each of these needs of a stakeholder are Objectives which are met by the system. Due to the cybernetic nature of all systems, the interconnectedness makes these objectives dependent on each other. Transitivity exists in their realization which can be defined variously. A natural hierarchy emerges in the realization of these objectives. This property helps us in the execution of these Objectives by defining natural sequence of realizing them. Definition of these objectives is thus an important skill and so determines the effectiveness of the strategy. In the instance of XC, a total of 40 Objectives were identified. These objectives determine the business processes. In the case in instance, these were used to guide the overall strategy development for viability of XC and its implementation
Margins critical to the operation influence the Cash available for Purchase of ECs. This determines the Purchase quantity of EC s which in turn determines the other costs like distribution, transport etc. Consequently the Margin resulting for XC is determined. Purchase Quantity and Cash Available determine the Margins and hence are critical to the Viability of the EC Operation Margins critical to the operation influence the Cash available for Purchase of ECs. This determines the Credit Period that is suitable for the deal. Accordingly the (Open) Market Prices and Purchase quantity of EC s will be fixed. The (Actual) EC Prices will be determined and consequently Other costs. Thus the Margins on the transactions will be influenced.
E. Understanding Variety
Cash availability, Credit Period and Market Prices together determine the Quantity of ECs that XC can buy. Accordingly the dependent parameters like other costs will drive the margins. This is a negative loop and as margins or cash availability falls, the quantity that can be bought will fall, and hence Margins will come down. This way the business will fall.
Organizations at the operation levels are essentially dealing with other entities, both internal to the organization and in the external eco system. Typically, when more than one entity interacts with other entities, one or more will control or guide and the others will follow the controller. Such interactions will be healthy only when there is stability between these entities. Hence from this perspective, making the transactions stable can be managed by ensuring that the Variety of the interacting entities are matched, either by enhancement or by attenuation of one or the other.
Based on the understanding given by the above Table 1 and similar analysis of the Influence diagram in Figure 3 below, some of the important factors that direct the solution definition are: 1.
Parameters 1, 5, 6 and 10 are focus areas and these are also difficult to change. Hence interventions to manage them needs attention
2.
Credit period is one of the critically important parameter for Viability of operations
Variety in Systems Thinking “refers to the multiple states in which a system can exist”. For instance “A switch has a variety of two (on and off) a child has a variety which is enormous [13] for any two interacting systems, the condition for stability would be when these varieties match. From this idea of matching variety between two interacting systems, we derive
319
the interventions that are required to make the interaction stable. Ashby’s says “only Variety can destroy variety [13].
To attenuate XC’s variety, we bring in Attenuators for the Market and Amplifiers for XC. This will thus be the strategy to make the operation viable. Accordingly, XC introduced a Consultant who could facilitate the Purchase the ECs at prices that will lead to good margins needed to make XC viable, given the low prices which the Tribal Buyers of ECs can afford. The Consultant had the challenge of Buying ECs in the quantity and Payment terms that was accept able to the market and sell them in the terms, quantity and Price that XC can afford.
“The term Variety was introduced by W. Ross Ashby to denote the count of the total number of states of a system. The condition for dynamic stability under perturbation (or input) was described by his Law of Requisite Variety. Ashby says [13] “The definition of variety is the number of states in which a system can exist. Variety is used to measure complexity.. If a system is complicated it has a larger variety. If it is simple the variety is low. “Managerial operational and environmental varieties should be designed to balance, with minimal damage to people and to cost.” [14]
This was possible from a study of the Markets. Large outlets called ‘Mandis’, directly buy the agricultural produce from the farmers when crops are harvested and sell them in large quantity to Wholesale buyers. Here purchase prices would be very low; nearly half of the open market, but quantity to be bought is large. For XC this quantities was as large as their yearly sale quantity. This was good to attenuate the variety of the Open market to a large extent. Payment terms and Volume of Payment were still difficult and so we need some more matching measures. From the variety matching perspective the approach looks as follows.
The instance of procurement of ECs by XC from the open market is typically a case of variety matching. Here the Purchase function of XC which is buying ECs from the market is faced with the problem that it can pay low prices to the seller of the goods in the Open Market. The Seller on the other hand can sell only at high prices for the quantity and credit that XC is able to pay. This mismatch is leading to the Purchase cost of ECs being high and therefore margin resulting being low. This is making the business unviable.
Variety Matching was used to evolve the strategy and verify it as shown in Figure 4 below
To solve this problem, we applied Ashby’s law of Requisite variety as follows: The Table 2 below details the Variety that is exhibited by the two interacting functions – XCs Purchase function and the EC Selling Function in the Market. The market variety is thus more than the variety of XC’s. Therefore XC finds it difficult to make its operation viable. TABLE 2 XC’s Purchase Function Cannot pay high price [since selling prices are low] Multiple entities influencing decision Can order only on lowest price
Payment terms are fixed fund flow not in control Quality important but ability to pay limited Others – Additional – Distribution, Remote, Markets, Storage management Total Variety
Variety of XC’s 1
Variety of Market 2
Market Conditions
3
Payment pricing: 1.
Quantity sensitive Pricing: 1. More qty less price 2. Less qty higher price
1 1
terms
Figure 4 . Strategy using Variety matching
sensitive
Against delivery low price 2. Late delivery high price 3. Pay when able very high price Flexible Payments- more than 3 modes Need promised funds on time
1
3
0
1
1
>3
3 or more qualities for agreed payment
1
>3
Delivery Modes – all at one time, supply when able- tied to pricing and long term deals
6
>15
F. Developing a Strategy for XC’s Operations Viability: The strategy driven by the formula 6+X = 15-Y was developed as : An agent or Consultant who could act as the interface between the Mandi’s (Markets) and XC, and absorb the higher variety of the Open Market and amplify the variety of XC’s Operations was identified. The features of the Mandi – selling large quantities only, Price being lowest compared to any other source, cost of transport and shipping were included in their selling price to XC. This part of the Open Market was chosen as the source of supply of ECs. On the other hand the variety of XC was amplified by leveraging the strengths of XC – Having the support of the Government and consequently having the creditworthiness, Assurance of usage of large quantity due its Buyer base, Having people who could run Purchase operations ( but not aware of Mandi management) The Consultant due to his ability to manage Mandi’s could manage the Purchase from
320
ACKNOWLEDGMENT
Mandi’s and train people. His not having the financial muscle was overcome by XC who gave the necessary Financial Guarantees and organized bank support by way of Letters of Credit to the Mandi stakeholders.
The authors would like to thank Tata Consultancy Services for providing the opportunity to learn, develop and apply the Multi modeling approach across several consulting assignments. Prof. P. N. Murthy for introducing them to the field of System Science, MGPL Narayana and other colleagues in the BSCC group for the feedback and encouragement provided.
Thus the strategy was to buy ECs from Mandi’s at lower prices and sell them at regular prices, thereby making the margins good and the operations viable. Other Systems Tools like Interpretive Structural Model (ISM) for Implementation Plan, Goals, Objectives, Targets, Outcomes (GOTO ) matrix for monitoring and tracking health of the operations were also used in this instance. The Strategy is captured in Figure 5 below.
REFERENCES [1]
Prigogine, I. and Stengers, I., Order out of Chaos, Flamingo, London, 1990. [2] Murthy, P.N., “Complex Societal Problem Solving – A Possible set Methodical Criteria”, Systems Research & Behavioral Sciences, 2001 [3] Rita Gunther McGrath, Gökçe Sargut, Learning to Live with Complexity, Harvard Business Review, september 2011 [4] P.N.Murthy, Systems Practice in Consulting, Systems Practice, Vol. 7 No. 4, 1994 [5] Jamshid Gharajedaghi, ‘Systems Thinking: Managing Chaos and Complexity: A Platform for Designing Business Architecture, Third Edition’,chapter 2- Morgan Kaufmann Publishers © 2011 [6] Beer, Stafford, The heart of Enterprise, London, John Wiley & Sons Ltd, 1979. [7] James Grier Miller, Living systems. New York: Mc Graw-Hill, 1978 [8] K. Sasidhar & P.N.Murthy, Complex problems and Qualtative Modelling. ,” unpublished. [9] Padmalata Nistala , Supriya Kummamuru et al An approach to understand and elicit requirements using systemic models: Ensuring a connect from problem context to requirements Procedia Computer Scxience 2013, [10] Sudheer V, P.N.Murthy, Multi modeling approach to Enterprise Analysis & Modeling July 1999 Unpublished [11] Sekhar Vadari, Supriya Kummamuru, A Cybernetic Learning Framework to Improve Quality of Higher Education, , Facets for Quality in Higher Education,pp 88-105, Macmillan Publishers India Ltd., 2011 [12] Andrew L. Delbeq, Andrew H. Van de Ven, David H. Gustafson, Group Techniques for Program Planning, Scott, Foresman and Co. Illinois, 1975 [13] W. Ross Ashby, An Introduction to Cybernetics, Chapman & Hall, 1956 Staffford Beer, David Whittaker, Think Before you Think: Social Complexity and Knowledge of Knowing, Wavestone Press 2009
Figure 5 Variety Matching Strategy for XCs operation
IV.
CONCLUSION
Problem Solving using Systems Thinking can be done in many ways. However, simple ways for this come from the idea that any problem is multidimensional in nature and is best understood by the use and application of multiple models for addressing various perspectives or aspects of the problem. This is broadly known as Multi modeling and has been usefully deployed in many instances to solve problems in differing applications. This paper has articulated the overall concepts in the models and some of the Systemic models that are relevant to traditionally common occurrences in industry. Among the many solutions developed by the authors and their employers, one has been chosen to illustrate the methodology and the process of strategy development.
321
