Toolkit: Complex Systems Toolkit.

Author: Milan Liu, Ph.D. Candidate (Cranfield University); Dr. Lampros Litos (Cranfield University). 

Topic: Towards circular economy: development of systems-based interventions in complex systems.

Title: Improving metal recycling and recycled content intake.

Resource type: Guidance article.

Relevant disciplines: Any; Production and manufacturing engineering.

Keywords: Recycled materials; Circular economy; Socio-technical systems; Waste management; Life cycle; Sustainability.

Licensing: This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License. 

Downloads: A PDF of this resource will be available soon.

 

Who is this article for?: This article should be read by educators at all levels of higher education looking to highlight the connection between complex systems and sustainability within engineering learning. 

Related INCOSE Competencies: Toolkit resources are designed to be applicable to any engineering discipline, but educators might find it useful to understand their alignment to competencies outlined by the International Council on Systems Engineering (INCOSE). The INCOSE Competency Framework provides a set of 37 competencies for Systems Engineering within a tailorable framework that provides guidance for practitioners and stakeholders to identify knowledge, skills, abilities and behaviours crucial to Systems Engineering effectiveness.  A free spreadsheet version of the framework can be downloaded.

This resource relates to the Systems Thinking, Life Cycles, Capability Engineering, Systems Modelling and Analysis, and Design INCOSE competencies.

AHEP mapping: This resource addresses several of the themes from the UK’s Accreditation of Higher Education Programmes fourth edition (AHEP4): Analytical Tools and Techniques (critical to the ability to model and solve problems), and Integrated / Systems Approach (essential to the solution of broadly-defined problems). In addition, this resource addresses AHEP themes of Materials, equipment, technologies and processes, and Sustainability.  

 

Learning and teaching resources:

Resource  Type  Best for  Quick classroom use  URL 
Insight Maker  Web-based modelling tool  Building stock-and-flow models and simple simulations  Convert the aluminium CLD into stocks/flows and run a scenario  https://insightmaker.com 
Loopy  Interactive causal-loop diagram app  Fast, visual CLDs and in-class demonstration of loop behaviour  Live demo of reinforcing vs balancing loops; students toggle link polarities  https://ncase.me/loopy 
Vensim PLE  Free desktop system-dynamics software  Introductory quantitative modelling and sensitivity runs  Short lab: implement simplified aluminium-recycling model and compare policy scenarios  https://vensim.com/free-download/ 
Leverage Points (Meadows)  Concept primer on leverage points  Framing where to intervene in systems  Assign as required reading; students map which leverage points the CLD targets  https://donellameadows.org/archives/leverages-points-places-to-intervene-in-a-system/ 
MIT  System Dynamics materials  Course notes and lecture videos  Structured curriculum and worked examples for deeper study  Use selected lectures and problem sets for follow-up or flipped classroom  https://ocw.mit.edu/courses/15-871-introduction-to-system-dynamics-fall-2013/  

 

Premise:

Several sustainability challenges, such as transitioning to a circular economy, are embedded in complex socio-technical systems. A circular economy is an economic model that replaces the linear take-make-dispose pattern with systems that keep materials and products in use for longer through designing for durability, reuse, remanufacturing, and recycling, while minimising waste and regenerating natural systems (Rizos, Tuokko, and Behrens, 2017).   

Complex systems like these exhibit feedback loops, delays, non-linear change, path dependence and emergent behaviour ​(Sterman, 2000; Meadows, 2008)​. This article introduces the idea of systems-based interventions using the example of aluminium recycling systems. It is designed for engineering educators who plan to provide learners with a baseline understanding of complexity and practical entry points for designing and developing and evaluating interventions that can move a system towards sustainability. 

 

Complexity of aluminium recycling systems:

Aluminium is infinitely recyclable, yet achieving truly closed material loops at scale remains a challenge. Most of today’s recycling occurs in situations where post-consumer scrap is collected from a wide variety of end-of-life products and the boundaries of the recycling system are difficult to define and control. This creates high variability in both the composition and the quality of recovered aluminium, since different products contain different alloys and levels of contamination ​(IRT M2P, 2023)​. At the same time, the volume of available scrap is difficult to predict, as it depends on product lifespans and consumer behaviour. These fluctuations make it harder for producers to plan and optimise secondary aluminium output, particularly when industries rely on consistent standards or just-in-time manufacturing. 

The recycling system is also shaped by broader economic and regulatory forces. On the one hand, demand for low-carbon materials and the cost advantage of recycled over primary aluminium are powerful drivers of growth. On the other hand, the system faces constraints from volatile scrap prices and shifting global trade dynamics, such as U.S. tariffs on aluminium imports. Meanwhile, new policy instruments are adding further complexity. The EU’s Carbon Border Adjustment Mechanism (CBAM) is set to reshape trade flows and investment patterns, while the forthcoming Digital Product Passport (DPP) will transform how information is shared across the value chain. Together, these forces influence technologies, markets and business models, underscoring the dynamic and interconnected nature of aluminium recycling. 

These interconnected factors highlight aluminium recycling as a complex socio-technical system, in which technological capabilities, market incentives, policy frameworks, and global trade are deeply interconnected. For educators, this makes aluminium an effective example for teaching students how multiple forces interact to create both opportunities and challenges for sustainable engineering. 

 

Intervention from systems perspective:

System Dynamics (SD), first formalised by ​Forrester (1968)​, has proven to be a highly valuable approach for understanding and managing complex resource and recovery systems. SD is an interdisciplinary approach, drawing on insights from psychology, organisational theory, economics, and related fields ​(Sterman, 2000)​. More supporting information about SD pedagogical tools and techniques can be found through the System Dynamics Society and Insight Maker. 

From a systems perspective, interventions are not isolated events but strategic effort to influence system behaviour by targeting its structure and dynamics. A key concept here is leverage points – places within a complex system where small changes can lead to significant, systemic effects ​(Meadows, 1999)​. ​Meadows​ identified twelve types of leverage points, ranging from adjusting parameters to transforming the system’s underlying goals and paradigms, proving a conceptual framework for identifying impactful intervention. 

Figure 1. Donella Meadows’ leverage points (Source: based on Meadows (1999); credit: UNDP/Carlotta Cataldi; reproduced from ​Bovarnick and Cooper (2021)​) 

 

Exploration of potential leverage points: 

System Dynamics (SD) tools such as Causal Loop Diagrams (CLDs) can help explore leverage points. CLDs can help visualise main components of a system and their interdependencies, making complex dynamics easier to understand. Besides, the process of building a CLD or more computational SD model encourages practitioners to clarify system boundaries, relationships, and drivers, laying the foundation for identifying leverage points. 

For example, a CLD of aluminium recycling might capture how classification and sorting processes influence scrap quality, which then affects remelting efficiency and ultimately market uptake of recycled alloys (see Figure 2 below). 

 

Figure 2. The causal loop diagram for auto aluminium recycling (Liu et al., 2025) 

By tracing these circular cause-and-effect relationships, learners can see where interventions may ripple through the system. Highlighting reinforcing loops, balancing loops, and delays also shows why some interventions produce limited short-term results but more substantial long-term effects. 

Leverage points can also be examined through the lens of information, rules, and goals. Improved information flows, such as those enabled by the Digital Product Passport, could reshape how scrap is sorted and valued. Rules, such as alloy specifications or trade tariffs, determine what types of recycled material can enter the market. At a deeper level, the goals of the system, whether to maximise throughput or to retain material value, fundamentally shape behaviour. Here too, CLDs are valuable because they allow users to visualise how changes to information, rules, or goals can shift system dynamics, providing a clearer picture of where interventions might be most effective. 

 

Implication for educators: 

This article equips educators with a focused, practical pathway to teach systems thinking through the example of aluminium recycling. Students can gain both conceptual understanding and hands-on skills to map feedback loops, identify delays, and design interventions that account for short-term trade-offs and long-term system behaviour. Teaching a single clear CLD followed by one modelling or scenario activity produces measurable learning gains while keeping the task accessible for beginners. 

 

Educational approach: 

 

Potential related learning outcomes within this topic: 

 

Further resources: 

 

References 

 

Any views, thoughts, and opinions expressed herein are solely that of the author(s) and do not necessarily reflect the views, opinions, policies, or position of the Engineering Professors’ Council or the Toolkit sponsors and supporters. 

Toolkit: Complex Systems Toolkit.

Author: Dr. Rhythima Shinde (KLH Sustainability).

Topic: Applying Cynefin framework for climate resilience.  

Title: Managing floods in urban infrastructure.

Resource type: Teaching – Case study.

Relevant disciplines: Civil engineering; Environmental engineering; General engineering.

Keywords: Systems thinking; Climate change; Sustainability; Risk; Decision-making; Problem-solving; Disaster mitigation.

Licensing: This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License. 

Related INCOSE Competencies: Toolkit resources are designed to be applicable to any engineering discipline, but educators might find it useful to understand their alignment to competencies outlined by the International Council on Systems Engineering (INCOSE). The INCOSE Competency Framework provides a set of 37 competencies for Systems Engineering within a tailorable framework that provides guidance for practitioners and stakeholders to identify knowledge, skills, abilities and behaviours crucial to Systems Engineering effectiveness.  A free spreadsheet version of the framework can be downloaded.

This resource relates to the Systems Thinking, Requirements Definition, Communication, Design For, and Critical Thinking INCOSE Competencies. 

AHEP4 mapping: This resource addresses several of the themes from the UK’s Accreditation of Higher Education Programmes fourth edition (AHEP4):  Analytical Tools and Techniques (critical to the ability to model and solve problems), and Integrated / Systems Approach (essential to the solution of broadly-defined problems). In addition, this resource addresses the themes of Sustainability and Communication. 

Educational level: Beginner; intermediate.

 

Acknowledgement: 

The case study underpinning this teaching activity was developed by Prof. Kristen MacAskill (University of Cambridge). The Module was ​first ​developed and implemented in teaching by TEDI- London, led ​by a team​ of learning technologists, Ellie Bates, Laurence Chater, Pratishtha Poudel, and academic member, Rhythima Shinde. This work was carried out in collaboration with the Royal Academy of Engineering through its Engineering X programme — a global partnership that supports safer, more sustainable engineering education and practice worldwide. With critical support from Professor Kristen MacAskill and involvement of Ana Andrade and Hazel Ingham, Aisha Seif Salim. This was a collective effort involving many individuals across TEDI-London and RAEng (advisors and reviewers), and while we cannot name everyone here, we are deeply grateful for all the contributions that made this module possible. 

 

Learning and teaching notes: 

This case study introduces a structured, systems-thinking–based teaching resource. It provides educators with tools and frameworks—such as the Cynefin framework and stakeholder mapping—to analyse and interpret complex socio-technical challenges. By exploring the case of the Queensland, Australia floods, it demonstrates how engineering decisions evolve within interconnected technical and social systems, helping students link theory with practice. 

The Cynefin framework (Nachbagauer, 2021; Snowden, 2002), helps decision-makers distinguish between different types of problem contexts—simple, complicated, complex, chaotic, and disordered. In an engineering context, this framework guides learners to recognise when traditional linear methods work (for simple or complicated problems) and when adaptive, experimental approaches are required (for complex or chaotic systems). 

Within this teaching activity, Cynefin is used to help students understand how resilience strategies evolve when facing uncertainty, incomplete information, and changing stakeholder dynamics. By mapping case study events to the Cynefin domains, learners gain a structured way to navigate uncertainty and identify appropriate modes of action. 

This case study activity assumes basic familiarity with systems concepts and builds on this foundation with deeper application to real-world socio-technical challenges.  

 

Summary of context:

The activity focuses on a case study of 2010–2011 floods in Queensland, Australia, which caused extensive damage to urban infrastructure. The Queensland Reconstruction Authority (QRA) initially directed resources to short-term asset repairs but subsequently shifted towards long-term resilience planning, hazard management, and community-centred approaches. 

The case resonates with global engineering challenges, such as flood, fire, and storm resilience, and can be easily adapted to local contexts. This case therefore connects systems thinking theory directly to engineering and governance decisions, illustrating how frameworks like Cynefin can support engineers in navigating uncertainty across technical and institutional domains. 

 

Learning objectives:

Aligned with AHEP4 (Engineering Council, 2020) – Outcomes 6, 10, and 16 on systems approaches, sustainability, and risk – this activity emphasises systems thinking, stakeholder engagement, problem definition, and decision-making under uncertainty. 

This teaching activity introduces learners to the principles and practice of systems thinking by embedding a real-world case study into engineering education (Godfrey et al., 2014; Monat et al.,2022). The objectives are to: 

 

Teachers have the opportunity to: 

 

Downloads: 

 

Learning and teaching resources:

 

Time required: 

The teaching activity is designed for 4–6 hours of structured learning, delivered across three modules: 

1. Context (1–2 hours) 

2. Analysis and insights (1–2 hours) 

3. Discussion and transferable learning (1–2 hours) 

 

Materials required:

1. Open access online learning platform: Engineering for a complex world

This dedicated platform hosts the interactive modules designed for this teaching activity. Students progress through three modules — Context, Analysis and Insights, and Discussion and Transferable Learning. Each module includes animations, narrative-driven content, scenario prompts, and interactive tasks. The platform ensures flexibility: it can be used in fully online, hybrid, or face-to-face settings. All necessary digital assets (readings, maps, videos, and quizzes) are embedded, so learners have a “one-stop” environment.

2. Case study pack: Queensland Reconstruction Authority flood response

The core teaching narrative is anchored in this Engineering X case study. It documents the evolution of the Queensland Reconstruction Authority (QRA) from a short-term flood recovery body to a long-term resilience institution. This resource provides students with authentic socio-technical detail — including stakeholder conflicts, institutional learning, and systemic barriers — which they then interrogate using systems thinking frameworks.

3. Facilitator’s guide: (Appendix A)

This guide equips educators to deliver the course consistently and effectively. It includes:

4. Timeline touchpoints: (Appendix B)

This resource provides a suggested delivery schedule for facilitators. It maps when live sessions, asynchronous tasks, and group discussions should occur, ensuring students remain engaged over the course. It also indicates where key reflective points and assessments (both formative and summative) can be integrated.

5. Pre- and post-module assessment form: (Appendix C)

This tool evaluates students’ systems thinking learning outcomes. It includes:

The form provides both quantitative data (Likert scales) and qualitative insights (open-ended reflections), enabling robust evaluation of teaching impact. 

 

Assessment:

 

Narrative of the case:

Learners are introduced to the case via a fictional guide, “Bernice,” who frames the scenario and supports navigation through the material. Students work through three stages that progressively apply the Cynefin framework and other systems tools to understand how resilience emerges through evolving governance and engineering responses: 

1. Context module: 

2. Analysis & insights module: 

3. Discussion & transfer learning module: 

 

Interactive learning design:

The teaching activity integrates multiple interactive elements to immerse students in systems thinking: 

 

Why this approach adds value: 

Although rooted in social-technical interactions, the activity explicitly connects systems thinking to core engineering design competencies—problem framing, stakeholder analysis, and iterative solution development under uncertainty 

 

Guided questions and activities: 

Facilitators can use these prompts to stimulate inquiry and structured reflection: 

 

Opportunities for extension: 

In addition to the Queensland floods and Sakura Cove examples, educators may draw parallels with urban heat planning in London, wildfire adaptation in Australia, or storm resilience in the Netherlands. These comparative cases allow learners to generalise systems insights beyond one event or geography. 

The activity is designed to be scalable and adaptable: 

This flexibility allows educators to tailor the activity to their students’ level of expertise, institutional context, and disciplinary focus. 

 

References:

 

Any views, thoughts, and opinions expressed herein are solely that of the author(s) and do not necessarily reflect the views, opinions, policies, or position of the Engineering Professors’ Council or the Toolkit sponsors and supporters.  

Toolkit: Complex Systems Toolkit.

Author: Dr. Stuart Grey, SFHEA (University of Glasgow).

Topic: Student created interactive simulation for complex sociotechnical systems.

Title: LLM-driven interactive simulation for complex sociotechnical systems.

Resource type: Teaching activity.

Relevant disciplines: Any.

Keywords: Artificial Intelligence; Large Language Model; Sociotechnical systems; Ethics; Modelling or simulation; Emergence.

Licensing: This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License. It is based upon the author’s article “Enhancing Ethical Reasoning in Engineering Education through Student-Created Interactive Ethical Scenarios Using Generative AI,” 2025 IEEE Global Engineering Education Conference (EDUCON), London, United Kingdom, 2025, pp. 1-5, doi: 10.1109/EDUCON62633.2025.11016531. 

Downloads: 

Related INCOSE Competencies: Toolkit resources are designed to be applicable to any engineering discipline, but educators might find it useful to understand their alignment to competencies outlined by the International Council on Systems Engineering (INCOSE). The INCOSE Competency Framework provides a set of 37 competencies for Systems Engineering within a tailorable framework that provides guidance for practitioners and stakeholders to identify knowledge, skills, abilities and behaviours crucial to Systems Engineering effectiveness.  A free spreadsheet version of the framework can be downloaded.

This resource relates to the Systems Thinking, Life Cycle, Configuration Management, Requirements Definition, Verification, and Validation INCOSE Competencies. 

AHEP mapping: This resource addresses several of the themes from the UK’s Accreditation of Higher Education Programmes fourth edition (AHEP4):  Analytical Tools and Techniques (critical to the ability to model and solve problems), and Integrated / Systems Approach (essential to the solution of broadly-defined problems). In addition, this resource addresses AHEP themes of Ethics and Communication. 

Education level: Intermediate.

 

Learners have the opportunity to: 

Teachers have the opportunity to: 

 

Overview:

This resource enables engineering students to create, run, and debug a text‑based, interactive simulation of a complex sociotechnical system using a Large Language Model (LLM). It is intentionally flexible and may be delivered as a multi‑session studio activity (including assessment) or used solely as a compact assessment.

  

Purpose and use:

In both modes, students design a robust text prompt, test it with a user, document changes, and submit auditable artefacts that evidence learning. The key activity is interrogating their own thinking on how complex systems should be modelled by making judgements as to how their game does and does not capture the system dynamics. 

 

Why and how: 

The approach aims to give students hands-on experience in putting systems thinking into practice. Concepts such as stakeholders, feedback loops, delays, uncertainty, and emergent behaviour can be implemented and interrogated without heavy tooling.  

The submission is a text LLM prompt with tracked changes, which allows students to demonstrate system design and debugging, produce transparent process evidence, and scale to large cohorts with minimal infrastructure. 

 

Delivery options at a glance:

Audience Undergraduate Years 2–4 and taught MSc, any engineering discipline 
Modes Studio activity (3–5×2 h + independent study) or Assessment‑only (prompt‑only; 1–2×2 h + 4–6 h)
Teams 3–4 students (solo permitted for assessment‑only) 
Assessment Portfolio (studio) or prompt‑plus‑change‑log (assessment‑only) 
Platforms Institutional Copilot licences successful; encourage exploration of free tools (students record model/version)

 

Materials and software:

 

Delivery modes:

Mode A — Studio activity (3–5 sessions) 

Mode B — Assessment‑only (prompt‑only; 1–2 sessions) 

In both modes, module leaders may supply a predefined scenario(s) to standardise scope and simplify marking. A ready‑to‑use example is provided in Appendix C. 

 

Assessment:

Studio portfolio — rubric (suggested weighting):

Criterion  A  B  C  D–E  % 
Complexity modelling  Clear boundary; rich stakeholders; ≥4 correct loops; delays explicit; coherent KPIs  Mostly sound  Basic map  Superficial  25 
Simulation design and prompt quality  Consistent state logic; visible feedbacks/delays; non‑linearity; negative choices allowed with consequences; clear commands  Mostly coherent  Playable but brittle  Confusing/linear  25 
Debugging evidence  Systematic play‑tests; clear issue → fix → re‑test artefacts  Some iteration  Minimal  None  20 
Insight and reflection  Deep analysis of emergence, trade‑offs, equity, uncertainty, and LLM limits  Good  Descriptive  Vague  20 
Communication and referencing  Clear, concise, correct Harvard referencing  Minor issues  Adequate  Disorganised  10 

 

Assessment‑only (prompt‑only) — compact rubric: 

 

Scenario options: 

Students may propose their own topic or the module leader may supply a predefined scenario. Options suited to UK engineering contexts include: 

 

Appendix A — Prompt template (simulation + debug‑ready): 

Title: Complex Systems Simulator — [Scenario] 

Purpose: Run a turn‑based interactive simulation of a complex sociotechnical system. Track named state variables, apply feedback and delays, and let the player’s decisions drive non‑linear outcomes. 

Setup: 

  1) Offer three roles (distinct authority/constraints). 

  2) Introduce 3–5 NPCs with clear goals and plausible interventions. 

  3) Show a dashboard of [STATE_VARIABLES] each turn with short context. 

State rules: 

Commands: status, talk [npc], inspect [asset], implement [policy], pilot [intervention], advance time, review log. 

Debug commands (for testing): trace on/off (print update logic), why (state which loops/delays drove the change), show variables (print current state table), revert (roll back one turn), reseed (slight exogenous shock). 

Realism and ethics: Allow all plausible actions and report consequences neutrally. If unsafe in the real world, refuse, propose safer alternatives, and continue with plausible systemic effects. 

LLM pitfalls to avoid: Do not invent new variables; ask clarifying questions rather than guessing; keep outputs concise; summarise trajectory every five turns. 

Begin: Greet the player, state the scenario, ask for a role, and wait. 

 

Appendix B — Debugging and play‑test checklist: 

Functional coherence 

Robustness 

User experience and clarity 

Report 

 

Appendix C — Predefined scenario (Urban Heatwave Response, UK city): 

Boundary: One UK local authority area during the July–August heatwave period. Focus on public health, energy demand, and community resilience. 

Roles: (1) Local Authority Resilience Lead; (2) NHS Trust Capacity Manager; (3) Distribution Network Operator (DNO) Duty Engineer. 

Stakeholders: Residents (with a focus on vulnerable groups), care homes, schools, SMEs, DNO, local NHS Trust, emergency services, voluntary/community groups, Met Office (for alerts), and local media. 

State variables (examples): Heat‑health alert level (0–4); Emergency Department occupancy (%); Electricity demand/capacity (% of peak); Indoor temperature exceedance hours (hrs > 27 °C); Public trust (0–100); Budget (£); Equity index (0–100). 

Events/shocks: Red heat alert; substation fault; procurement delay; misinformation spike on social media; transport disruption; community centre cooling failure. 

KPIs and stop conditions: Heat‑related admissions; unserved energy; cost variance; equity gap across wards. Stop if alert level 4 persists >3 days, budget overspends >10%, or trust <25. 

Notes for assessors: Using a standard, predefined scenario simplifies marking and ensures comparable complexity across teams, while still allowing for diverse strategies and outcomes. 

 

Any views, thoughts, and opinions expressed herein are solely that of the author(s) and do not necessarily reflect the views, opinions, policies, or position of the Engineering Professors’ Council or the Toolkit sponsors and supporters.  

Let us know what you think of our website