Comprehensive Study Notes: Systems Thinking and Engineering and Engineering and for Society Systems Thinking and Systems Analysis

Systems Thinking and Systems Analysis at TU/e

• Context of Systems Engineering: The concept of a "system" is a foundational pillar in engineering at Eindhoven University of Technology (TU/e). It is taught and practiced across various specialized groups, emphasizing that systems thinking adds a necessary layer to traditional "non-system" thinking.
• TU/e System Groups and Mission Areas:
  • Autonomous and Complex Systems: Science and technology for the next generation of autonomous systems.
  • Control System (CS): Dynamic modeling and model-based control of complex dynamic systems.
  • Design and Decision Support Systems: Planning and management of large-scale urban artifacts like cities, malls, and offices.
  • Electrical Energy Systems: Focus on sustainable energy advancement, from renewable generation to distributed power units.
  • Electronic Systems: Creating high-quality, cost-effective design trajectories with predictable properties like functionality, timing, and power dissipation.
  • Eindhoven Artificial Intelligence Systems Institute (EAISI): Collaborates across departments to create AI applications with real-world impact.
  • EIRES Systems Integration: Scientific discussions on the integration of various energy systems.
  • Embedded Control Systems Lab (ECS): Codesign and optimization for control implemented on resource-constrained platforms.
  • Engineering for Sustainable Energy Systems: Designing solutions for energy systems involving (electro)chemical transformation and multiphase transport.
  • Formal System Analysis: Techniques for modeling and analyzing concurrent systems using process algebra, model checking, and logics.
  • Information Systems (IE&IS and W&I): Studying tools to help organizations use information for operational decision-making.
  • Information Systems in the Built Environment (ISBE): Engineering and management of info systems embedded in the built environment.
  • Interconnected Resource-Aware Intelligent Systems: Addressing challenges in timing, dependability, and energy efficiency for distributed systems.
  • Microfluidics and Soft Matter Microsystems: Using microfluidic tools to study physical mechanisms of soft materials.
  • Microsystems: Inspired by biology, applied to understand health and disease processes.
  • Mobile Perception Systems Lab: AI for autonomous systems to anticipate events in dynamic environments.
  • Networked Embedded Systems Lab: Design and optimization of low-power networks for the Internet-of-Things (IoT).
  • Neuromorphic Edge Computing Systems Lab: Creating architectures for sustainable and emerging computing applications.
  • Signal Processing Systems: Algorithms and solutions for interpreting images, signals, and multivariate data.
  • Spatial-Temporal Systems for Control: Modeling multi-physics dynamic systems.
  • Systemic Change: Designing technology-enabled interventions at the community level to address societal challenges.
  • Terahertz Photonic Systems: Exploring waves for wireless communication, remote sensing, and material probing.

NASA Systems Engineering Framework

• NASA Definition of Systems Engineering: A methodical, multi-disciplinary approach to the design, realization, technical management, operations, and retirement of a system.
• Defining a System (NASA): A combination of elements that function together to produce the capability required to meet a need. It includes hardware, software, equipment, facilities, personnel, processes, and procedures.
• Value of Integration: The true value lies not in individual parts, but in how those parts interact through interconnection and integration.
• The Big Picture: Systems engineering focus on a holistic framework to manage complexity and enable informed technical decision-making.
• Art and Science: Systems engineering is both an art and a science, focused on meeting requirements within conflicting constraints (structural, electrical, mechanical, and human factors).
• Optimal Solutions: The approach seeks balance, ensuring no single subsystem is prioritized at the expense of others. It focuses on building the system right (meeting requirements) and building the right system (delivering operational outcomes).
• Role of the Systems Engineer:
  • Ensures technical requirements are met and sound practices are followed.
  • Coordinates the technical team and guides development from concept to deployment.
  • Key Responsibilities: Developing Concept of Operations (ConOps), system architecture, defining boundaries, conducting trade-off analyses, managing interfaces, and leading verification/validation.

Definitions and Components of a System

• Variability in Definitions: Philosophers and scientists define systems based on their focus (social, technical, or environmental).
  • Anderson & Johnson (1997): A group of interacting, interrelated, or interdependent components forming a complex and unified whole.
  • Kim (1999): Adds that the system must have a "specific purpose."
  • Meadows (2008): Focuses on three pillars: Elements, Interconnections, and a Function or Purpose.
  • Wikipedia: Mentions that a system acts according to a "set of rules" and is surrounded/influenced by an environment.
  • Parsons (2012): Defines a social system as a plurality of individual actors interacting to optimize gratification via culturally structured symbols.
  • Luhmann (2006): Abstractly defines a system as the "difference between system and environment."
• Essential Components (Meadows 2008):
  • Elements: Often the easiest to identify (visible/tangible). Can be material (roots of a tree) or intangible (academic standards of a university).
  • Interconnections: The relationships holding elements together. Can be physical flows (blood in a digestive system) or flows of information/signals (a consumer reacting to price changes).
  • Function or Purpose: Often the hardest to see. Deduced from behavior rather than stated goals. Example: A frog's purpose is catching flies, regardless of which way it turns.
• Hierarchy of Importance: Elements are often the least important in defining a system's character. Changing the players on a football team keeps it a football team. Changing the rules (interconnections) or the goal (purpose) radically transforms the system.

Key Characteristics of Systems

• 1. Holistic Integrity: A system's parts must all be present for it to carry out its purpose optimally. If removing a part doesn't change the function, it is likely just a "collection" (e.g., a bowl of nuts).
• 2. Specific Arrangement: Components must be organized purposefully. Random arrangements fail (e.g., accountants taking over marketing roles without training).
• 3. Nested Purpose: Systems exist within larger systems. You cannot divide a system and get two smaller versions of the same thing (e.g., splitting an elephant doesn't produce two small elephants).
• 4. Stability through Fluctuations: Systems maintain stability around a preferred range (homeostasis).
  • Example: Human body temperature stays near $98.6^{\circ}\,\text{F}$.
  • Example: Organizations maintain target profit margins through continuous feedback and adjustment.
• 5. Reliance on Feedback: Information is transmitted and returned to prompt behavioral adjustment. Feedback can be instantaneous (steering a car) or delayed (sun exposure in youth leading to skin damage decades later).

The Event-Pattern-Structure Pyramid

• Events (The Tip of the Iceberg): Immediate, visible occurrences (e.g., a machine breakdown). Solving at this level addresses symptoms only.
• Patterns: Recurring trends over time. Seeing how events connect (e.g., projects consistently going over budget).
• Structure (The Root Cause): The underlying arrangement of relationships and rules. This invisible framework drives patterns.
  • Example: Senior engineers leaving a company because of a policy change that increased administrative workload, creating a self-reinforcing cycle of stress and resignations.

Linearity vs. Non-Linearity

• Principles of Linear Systems:
  • Additivity: The net response to multiple stimuli is the sum of individual responses.
  • Homogeneity: Output is always proportional to input ($\text{input} \times 2 = \text{output} \times 2$).
• Limitations of Linear Models: They fail to capture feedback, scaling effects (economies of scale), or interactions (e.g., two drugs interacting in the body).
• Non-Linear Systems: Cause and effect are not proportional.
  • Emergence: Interactions create novel phenomena (e.g., Hydrogen + Oxygen = water with the property of "wetness").
  • Tipping Points: Soil erosion may have zero impact on crop yield until topsoil is worn to the root zone, at which point yields plummet.
• Exponential Growth and the Magic Number 70:
  • Exponential growth follows a J-shaped curve.
  • Rule of Thumb: Doubling Time $= \frac{70}{\text{Growth Percentage}}$.
  • Example: A $2\%$ growth rate leads to a doubling every $35$ years ($\frac{70}{2}$).
  • Example: Chinas $10\%$ annual growth led to doubling every $7$ years ($\frac{70}{10}$).
  • The Water Hyacinth Scenario: A lake of $10\,\text{km}$ diameter requires $8 \times 10^9$ plants to be covered. If the population doubles monthly, a lake only $0.2\%$ covered at month $24$ will be $13\%$ covered by month $30$ and completely covered by month $33$.

Paradigms: Analysis vs. Synthesis

• Analysis (Reductionism):
  • Traditional scientific method.
  • Process: Decompose the whole into parts, analyze parts in isolation, then recombine.
  • Implicit Belief: The whole is merely the sum of its parts.
  • Works best for systems with low interconnectivity.
• Synthesis (Holism/Systems Thinking):
  • Focuses on the context and relations within the whole.
  • Process: Identify the relevant system, understand how the whole functions, then examine how parts are interconnected to serve that whole.
  • Implicit Belief: The parts are explicable only by reference to the whole.

Epistemology and Mereology

• Realist vs. System/Constructivist Views:
  • Realist: The world is exactly as engineering laws (like Ohm's Law) describe it.
  • Constructivist: Science provides models of reality, not reality itself. Example: Ohm's Law ($V = I \times R$) is an empirical model that fails in scenarios like varistors or extremely strong electric fields.
• Assumptions Comparison:
  • Classical: Equivalence, Continuity (smooth changes), Mechanic (machine-like), Elementaristic (sum of parts).
  • Systems View: Bifurcations (tipping points), Phases (phase transitions), Self-interaction/Autonomous order, Macroscopic emergent qualities.
• Mereology: The theory of parthood relations ($\mu\epsilon\rho o\varsigma$ = part).
  • Parts can be attached (handle of a mug) or detached (remote control of a stereo).
  • Parts can be spatial (area of a room) or temporal (first act of a play).

Stocks and Flows

• Stock: An accumulation of material or info built up over time. It represents the "memory" of the system.
  • Examples: Water in a tub, population, bookstore inventory, self-confidence.
• Flow: The rate of change (input or output) per time unit that alters stocks.
  • Examples: Births, deaths, purchases, sales, deposits.
• Bathtub Archetype: Inflow comes from a Source, outflow goes to a Sink. The stock is the water in the tub.
• Forest System Example:
  • Inflow: Tree growth (governed by seed germination, climate, light).
  • Outflow: Tree death (parasites, pollution) or logging (industrial activity, market demand).
  • Feedback: As tree population drops, parasites might decrease, which eventually slows the death rate (self-regulation).
• Dynamic Sustainability Indicators:
  • Turnover Time: $\frac{\text{Stock Size}}{\text{Stock Change Rate}}$.
  • Coverage Time: $\frac{\text{Stock Size}}{\text{Drain on Flow}}$.
  • Harvest / Regeneration Index: Sustainable if $\le 1.0$.
  • Emission / Absorption Index: Sustainable if $\le 1.0$.

Feedback Loops: Mechanics of Dynamics

• Reinforcing Feedback (Positive Loop): Amplifies change and leads to exponential growth or collapse.
  • Diagram Polarity: $\text{A} \xrightarrow{+} \text{B} \xrightarrow{+} \text{A}$.
  • Examples: Population growth, disease spread, economics of scale (more sales leads to more revenue which leads to more scaling and lower costs).
• Balancing Feedback (Negative Loop): Opposes change and creates stability/equilibria.
  • Diagram Polarity: $\text{A} \xrightarrow{+} \text{B} \xrightarrow{-} \text{A}$.
  • Examples: Body temperature regulation (sweating/shivering), steering a car (correcting deviation), airplane autopilot.

Causal Loop Diagrams (CLDs)

• Purpose: Qualitative tool to represent mental models and feedback structures.
• Components:
  • Variables: Nodes representing system elements.
  • Arrows: Represent causal influence.
  • Polarity (+/-): $(+)$ means variables move in the same direction; $(-)$ means they move in opposite directions.
  • Loops: Labeled R (Reinforcing) or B (Balancing).
• Software: Vensim is a primary tool for building these diagrams.
• The Narrative Method: Start with a story (e.g., "More eggs lead to more chickens, but more chickens crossing the road leads to more deaths") to identify variables and links.

Case Study: VITO and Organizational Legitimacy

• Core Concepts: Legitimacy is the ability to defend decisions with justification. Organizations judged legitimate receive resources and "strategic freedom."
• VITO Innovation Journey (Deep Geothermal Energy - DGE):
  • Chapter 1 (Winning land): Won legitimacy with local communities via potential for jobs and sustainable energy.
  • Chapter 2 (Funding and Drilling): Won legitimacy with investors using a business model. Drilled 3 wells: 1 successful, 1 limited, 1 dry (catastrophic for the business case).
  • Chapter 3 (Pivoting to R&D): Re-legitimized by focusing on research. Operation began, followed by a seismic event (earthquake) in June 2019.
  • Chapter 4 (Crises and Recovery): Lost legitimacy due to risk. Regained it via public hearings, external experts, and focusing on climate goals. Plant restarted in 2021 after a $19$-month shutdown.
• The "Success Trap" in CLDs:
  • Loop R1 (Chasing Legitimacy): Successful strategy increases benefits, encouraging persistence in that strategy.
  • Loop B1 (Legitimacy Challenge Diversification): Increased activity draws broader interest, which introduces new, diverse criteria for legitimacy, making the old strategy obsolete.
  • Loop B2 (Succession Delay): Organizations with finite resources spend less on updating strategies while current ones work, creating a delay and eventually a legitimacy crisis.

Case Study: Plastic Pollution System Modeling

• Complexity: Plastic pollution is ubiquitous and irreversible. A workshop by Morasae et al. (2024) identified $18$ factors and $7$ feedback loops.
• Key Factors: Effective legislation, Funding, Public education, Behavior change, Innovation, and Waste management.
• Modeling Choice:
  • Use CLD to understand why things happen (qualitative).
  • Use SFD (Stock-Flow Diagram) to quantify how much and how fast (quantitative). Examples include modeling municipal plastic waste flow in Khulna City.
  • Methods like Life Cycle Analysis (LCA) and Material Flow Diagrams provide specific views on the global plastic cycle (measured in $\text{Mt/y}$, million metric tonnes per year).