Approach: capability, cycle and step explanations
The iteration cycle, worked through five real cases
The Capability Pipeline
About this resource · How to use it · Learning engineering connections
What is this resource?
Five worked narratives apply a seven-step capability cycle with continuous refinement across healthcare, government services, aviation, education and autonomous vehicles. Teaching interpretations accompany the source narratives; historical claims have not been independently rechecked here.
How should I use it?
- Pick a case and begin in Study mode to inspect the seven numbered steps and the continuing refinement reflection. Use the central Refine prompt to revisit any earlier assumption.
- Try Find the turn or Predict first before revealing the decision-specific explanation. Compare your reason with the presented evidence.
- Compare two cases and use the coupled-loops diagram to identify how a change to work conditions could change learner performance.
How does this connect to LENS and learning engineering?
Comparing contrasting cases supports conditional judgment: the same process can turn on different constraints. In LENS terms, map both the human capability and system conditions, then ask which measurement would reveal a failure. Locating a turning point in a worked case does not establish transfer to a new system.
Seven steps, one loop: understand → map → design → build → instrument → experiment → evaluate → understand. Refine sits at the center: use evidence to improve any part of the work throughout the cycle. Start with a challenge; develop a capability that acts on the problem. Explore the seven steps and a continuing refinement question across a hospital screening workflow, a federal insurance marketplace, an airliner, a math-tutoring program at scale, and a self-driving car. What changes case to case isn't the shape of the loop — it's which single step the story actually turns on.
The cycle in use
A challenge enters Understand through the dotted arrow. Follow Steps 1–7, then return from Evaluate to Understand. Refine throughout. Deployment connects the resulting capability back to the challenge through adoption and effective use.
The step that decides the story is rarely evaluate. It's usually earlier — design, build, or instrument — because those steps determine what evaluate is even able to see. A system instrumented to track only its own headline metric cannot evaluate a cost that metric doesn't measure.
Below, five cases — one apiece from government services, aviation, healthcare, education, and autonomous vehicles — are walked through all seven steps, with a continuing refinement reflection. Each carries one clearly marked point where a different upstream decision would plausibly have produced a different downstream result.
Learning guide. Open a step for why it matters, a decision cue and a question to apply to this case. Teaching commentary is an interpretation of the presented narrative, not new historical evidence. In Find the turn and Predict first, the decision-specific explanation appears only after your guess or reveal.
Pick a case
Same seven steps and continuous refinement, five different systems. Each card names the domain, the real frame, and how the case actually turned out. Three ways to work a case are below the picker — study it worked out, find where it turns yourself, or commit to a call before you're shown what happened.
Integrative Practice
Learners and systems develop together
The seven steps, with continuous refinement, support linked cycles of learner development and system evolution. Joint practice couples them at every step: people influence the system, and experience with the system shapes learning.
- Understand & Map — agree on the challenge
Bring learner goals, existing knowledge, system needs and operating conditions into the same picture. Learners help define success and influence choices.
Ask: whose goals and requirements shape the work?- Design & Build — create together
Co-design learning conditions and system behavior. Make roles, control and opportunities for creative partnership explicit, then test how they work together.
Ask: what should people learn, and what should the system enable?- Instrument & Experiment — test under representative conditions
Plan measurement of learning, system quality and joint performance. Deployment makes a system available; adoption, effective use and sustainable support require attention in real practice.
Ask: what can we observe, and what makes successful use possible?- Evaluate — return to Understand; refine both loops throughout
Examine learner knowledge, capability and flourishing alongside technical requirements and collective results. Joint performance alone does not establish independent learner competence.
Ask: what evidence calls for a change to teaching, the system, or the shared challenge?
Integrative Practice keeps the loops coherent. Connect each choice to its rationale, evidence needs and subsequent revisions. Use findings from authentic practice to revisit goals, resources, roles and learning conditions.
This is a conceptual view of how the seven-step cycle and continuous refinement fit together. Learner development, system evolution and collective capability each need their own evidence.