Systems Thinking for Design
Systems Theory and
Problem Structuring
Dr. Srikar A.
School of Interdisciplinary Design and
Innovation (SIDI)
Session outline
Introduction to Systems Theory
Problem Structuring using Concept
Maps
2
Source: Internet
• The problem of reductionist thinking … limited/narrow
Inter-disciplinary disciplinary view
approach to • The challenge of integrating disciplinary concepts …
incommensurability
problems • Dealing with socio-technical problems … in search of
common language
Fundamentals of Inter-disciplinary approach
INSIGHT
Problem Scientific
Vehicle Traffic situation situation Fluid Dynamics
ANALOGY Perception
Conceptual Conceptual
Model Model Kinematic model:
Flow-concentration curve
Deduction ISOMORPHISM
Induction Homomorphism
Abduction Rigorous Rigorous
Formulation Formulation
Kinematic shock-wave
Model based Systems Engg
Bio-inspired Design
Scientific Generalization
Source: Stafford Beer (1966), Decision and Model
Control
4
Advance
s in inter-
disciplina
ry
theories
The language of systems … complexity
• Systems, Cybernetics and Complexity as inter-disciplinary
concepts & languages
• System / Complexity as a key property of coherent,
dynamic, adaptive phenomenon
• Most socio-technical problem situations have good and
bad complexity
• Simplification should eliminate bad complexity, not good
complexity
• Welcome to complexity!
6
Variety as a measure of complexity
B B B
A A A
C
A C
G
B A G F
B D
D
F
A E E
A collection of partial similars A collection of dis-similars An assemblage of dis-similar
Source: Stafford Beer (1956), Decision and 𝑉𝑎𝑟𝑖𝑒𝑡𝑦 (𝑘)=𝑛=7 (coherence),
Control
7
Variety as a measure of complexity
B B
A A
0/1
C C
G G Intelligent,
adaptive, living
D D
systems need to
F F have a high degree
E E of complexity
A systematic assemblage of dis-similar A dynamic system
Source: Stafford Beer (1956), Decision and
Control
8
Why systems for design and innovation?
System: Design:
Innovation:
A pattern that is A pattern that is A pattern that has
coherent and has distinctive, yet
become an attractor –
emergent properties > contextual – engaging
dynamic & growing
than the sum of parts & empathetic
Patterns differ in terms of the degree of complexity (a function of N,K,C)
9
The three concepts deal with different levels of complexity
High complexity
Society/Ecosystem
Innovation
Organization
Weak Cause-Effect Design
Outcome based recognition Thinking
Product/System Complexity
Systems Theory
Engineering
Sub-system
Strong Cause-Effect Engineering
Component Input based recognition Discipline
Low complexity
1
Aspects to be
FORM FUNCTION BEHAVIOR STRUCTURE PROCESS E&M E&M
learnt
Function Behavior Structure Prototyping Mobilizing Leading
Social Paradox-
Form Team forming & Team Norming / Performing / External Beyond a team… Reflexive-
aspects storming / Iterative Interactive Integrative Negotiation / pitch network Leadership
Mnf Industry Focus Observe Product Understand Understand desired Understand Build prototypes Understand Understand ethical
(Space, Auto, M/c Sketches in these emerging needs in product behaviors structures in these relevant to these competition in issues in these
Tools, Proc Industries) verticals / history these verticals in these verticals verticals domains these verticals verticals
Why & How to Art-Part, narrative Simple -complex Technical-Social Creative Synthesis Product-Process
Tech-Design- Social-economic-
achieve Inter- writing & Sketching Systems as a ANT, Network of M/E/I Abductive (Design
Ma-nufacture) Business ecological
disciplinary Synthesis for thinking & com common language perspective, Semiot / Metaphor
Customer Stakeholder Customer / Customer/user
User reviews & Customer Value /
Focused Analysis / Need unstated needs / perceptions of form
usability Market Insight
Identification user contexts – affordances/signs
Product-Env Defining product Product level Concept eval, MVP,
Product-Enterprise
Product Level First Exterior form (Deriving product level behaviors / structural decisions prototyping choices
fit
level functions) attributes (Form-Beh-fn Fit) – COTS/AM/SRP
Understanding patterns
Unbundle and Functional arch to Strategy-Orgn-
that hold the whole Derive Functional Component /
reassemble … support desired
(System-Sub-system- Hierarchy technical arch & Bs Process/Fn
behaviors (smart)
Component-Part)
Likely evolution of Likely evolution of Process Business
Lifecycle perspective Likely evolution of
function-env over behaviors over time Quality plan
(Time) structure over time
time - Reliability
Concept sketch / 3D Ethnography / Rich Biomimicry / PUGH / TRIZ / MVP/ Porter’s f/w, Lean
Issue map / SNAC / CSH?
Techniques and Tools rendering / Physical Ps / VSM / Morphology / Config Mgt / Prog startup, Cost &
realization ISM / SysML–Fn SysML-Str Financial Models
Mgt, RP Tools
SysML- Be
Notebook used Core quality Working PoC –
Cloth on table House using
Point of Passage to represent
house to create space
using form & fn bamboo etc.
house with
bricks
Seed customer 1
Emergent Whole & Part (n)
• Anything is a whole which operates in quasi-independence of its environment.
Example, a rock, a mammal, an apple, a committee
• Parts are the immediate basic factors into which wholes can be analyzed, and
they are of a certain quantity… example, a rock analyses into parts as
crystals, an apple into such parts as skin, flesh and seeds
• Sub-parts are the second level of analysis. They are immediate factors into
which parts can be analyzed. Example, a crystal analyzes into molecules,
apples flesh into cells
• Wholes can be parts of still larger wholes, and sub-parts can be analyzed into
sub-sub-parts. Example, rocks are part of mountains, apples of the tree-
system, and molecules can be analyzed into atoms, and apple cells into
molecules
• Systems and Complexity differ in the way they view parts and wholes… in
complexity
Source:
Thinking
wholes are emergent properties, whereas in systems wholes are
Emery (1969), Systems
treated as a level
1
Types of relations (k)
• Transitive: Relation of two parts through a middle part
• Symmetry: Where interchange of parts does not involve any change in the
relation
• Additional types of relations can be a combination of the above. For example,
seriality is the relation which is transitive and asymmetrical
• Correlation: A relation between two series such that for every part in one
series there is a corresponding part in the other series
• Dependence: Relation in which the existence of one part is conditioned
by some other, example, limb of an animal is dependent upon the
circulatory system
Source: Emery (1969), Systems
Thinking
1
System = Function of Parts, relations, content/rules
• The rules or principles according to which all relations fall together into one
controlling order makes an organization or a system. The level of complexity
depends on the degrees of freedom of parts, relations and rules
• Systems and complexity theories differ in the way they treat the rules or
principles. Complexity assumes that rules of interaction between parts (local)
decides emergence … flight of geese, whereas systems looks for rules or
principles in the purpose of the whole (macro)
• Boundary,
• BoundaryEnvironment, Closed
can allow/prohibit and
exchange of Open
matter, Systems
energy, information with the
• environment Boundary judgments (what is inside is sacred, what is outside is
profane)
Source: Emery (1969), Systems
Thinking
1
A Systems View of Physical Sciences
PHYSICAL UNIVERSE
Relations among concepts (parts)
UNIFIED FIELD
COSMOLOGY
MASS HIERARCHY STATISTICAL STATE
More deterministic
SPACE-TIME ACTION UNCERTAINTY IRREVERSIBILITY
SPACE TIME ENERGY INFORMATION
CHANCE
Source: Seshagiri (1983), Fountain Heads of Science 1
A Systems View of Chemical Sciences
CHEMOSPHERE
Relations among concepts (parts)
UNIFIED FIELD CHEMICAL ECOLOGY
CHEMICAL CHEMICAL DIVERSITY
REACTIONS HIERARCHY
More deterministic
STRUCTURALITY SPECIFICITY VARIABILITY EVOLUTION
ELECTRON INVARIANCE TRANSPORT INFORMATION
CLOUD CHANCE
Source: Seshagiri (1983), Fountain Heads of Science
1
A Systems View of Biological Sciences
BIOSPHERE
Relations among concepts (parts)
INTEGRATION ECOLOGY
PATHWAYS DIFFERENTIATION DIVERSITY
More deterministic
STRUCTURALITY SPECIFICITY VARIABILITY EVOLUTION
BIO-PRECURSORS INVARIANCE TRANSPORT INFORMATION
CHANCE
Source: Seshagiri (1983), Fountain Heads of Science 1
Session outline
Introduction to Systems Theory
Problem Structuring using Concept Maps
1
Fuzzy front-end of product design
Deploy /
Discovery
Operate
This is the fuzzy front-end (FFE) of
product design (a chaotic phase
with no clear cause-effect rules)
Develop /
Diagnosis
Manufacture
Design
1
Data for discovery
All these influences are potential
sources of data for discovering new
Deploy /
Operate
Discovery product concepts
Usually involves people from
different divisions, with different
specializations and priorities
Develop /
Diagnosis
Manufacture
Important that whatever role you
chose, you are sensitive to problem
Design discovery
2
Data collection methods: should increase
complexity for better problem understanding
Qualitative & Quantitative Problem
data from primary
(observation, interview, Description
Images, Voice, Text,
group techniques) and
Data
Document
secondary sources
(research, reports, news
items)
(discovery
matrix)
Different Data Outcome of
Multiple sources
types Discovery
Tools like Nvivo support this entire process, largely in academic research
2
Discovery matrix: starting point for diagnosis
(qualitative research)
• Once we have represented a situation in the form Parts (Key 1 2 3 4 5 6
of parts and relations, we can ask different 1
Elements)
questions about the situation 2
• Is the representation holistic? Are we missing any 3
4
stakeholder perspectives? For instance, for the fuzzy 5
front-end of product design (conceptual design) the 6
stakeholders could be customers, competition,
academic research, shareholders, management,
engineering design, manufacturing
• Second, is there a pattern in this description? Is it
biased or focused only on a few aspects? What are the
critical points? Or patterns that cut across some
elements?
• Some of these patterns could be extremely beneficial
and represent the complex character of the
phenomenon, while others could be detrimental and
complicated
2
Diagnosis: Using complexity principles to extract a
more creative, holistic and rigorous problem
definition / product concept
Problem Description Pattern Extraction Problem
(Discovery Matrix) using Principles of Definition or
Complexity Requirement
2
How order can emerge out of chaos?
Initial Interest Discovery Diagnosis Design
80 inputs: general 238 inputs: product 825 inputs: product
20 inputs: nothing
perimeter architecture design
Source: [Link]
Does this follow the Golden Ratio?
2
• Identify the parts and relations from the
stateme
problem
nt
• Abstraction is key to identify parts…
times…
ask why at
5 least 10 elements
• Use a simple relation to start with – a “is
Exercise 4.1 to” to
connected
b
(30 min) • Use a matrix to depict parts and
10
relations (10 by
)
Example: Keywords from policy statement -“No Plan
At Present To Have All Electric Car Fleet By 2030”
26
Example: Keywords from opinion – “Electric
Vehicles In India; What The Future Holds”
27
Example: Keywords from a counter view: “Electric
vehicles don’t need a government push”
28
Discover
underlying
complexity 29