The cycle

Practice
Research
Foundation
Instruments
Practice

Practice generates real signals. Research forms questions from them. Foundation stabilises what holds. Instruments return it to action — and back into practice.

The cycle persists because practice keeps producing new boundary cases, unexpected signals, and revised contexts. It does not close.

How practice creates pressure

The system does not just produce findings. It generates pressure on them.

Discovery Loop
From observed signal to published structural claim
Signal
Case
Structural reading
Publication

A signal is raw contact — something that resists, doesn't fit, or recurs. A case examines it structurally. A structural reading turns the signal into a claim that can be discussed, challenged, and carried across contexts. Publication compresses that claim into transmissible form.

Publication exposes the claim to further evidence
Pressure Loop
Published claims remain open to boundary pressure and revision
New evidence
Boundary pressure
Revision
Stabilisation

Published findings remain open to new evidence, boundary cases, and stronger explanations. A published case is not only an archive entry — it is a standing claim whose scope can become more precise over time.

Publication does not freeze a finding. It operationalizes it.
Standard publication systems optimise for archival stability and local correctness. StratoAtlas treats publication as a point where a claim becomes available for continued testing, comparison, and refinement.

The system is designed so that revising a weak finding is healthier than silently preserving it.

How findings remain revisable

Controlled revision — not arbitrary revision
Revision remains possible, but not arbitrary.
Creates revision pressure
New evidence from a different context challenges the scope of the claim
A pattern that appeared structural proves narrower than originally described
Further examination reveals a false match or an unaccounted boundary condition
Accumulated edge cases create enough pressure to revisit the classification
Does not carry enough pressure
Repetition of the same claim without new evidence
A competing explanation without a clear way to test it
Informal reassessment without new structural evidence

Diagnostic series

Cases accumulate into series when the same structural class of contradiction recurs across a domain. Series are not settled conclusions — they are living findings whose scope can sharpen as new cases appear.

Domain series — cases from specific industries
Structural class — cross-domain mechanism
Extended domain — production failures across industries
Recurring finding — across series

The absent layer determines where the contradiction migrates.

In ERP, the system logic meets organizational logic and the gap encodes into the layer no one owns. In AI deployments, the model capability meets operational context and the gap encodes into evaluation architecture. In MedTech, the system logic meets care logic and the gap migrates to the boundary nobody was asked to design. In failure regimes, the corrective channel collapses — or was never orthogonal to what it was supposed to correct. Different domains, same structural mechanism. Practice is where this mechanism becomes visible before the damage becomes irreversible.

01

Observations

Structural findings captured early enough to matter. Each observation is a compressed signal from practice — named before it fully becomes a case.

Loading…
02

Cases

Situations where a contradiction, failure, or structural tension is examined in context. Cases are the primary unit of analysis in practice.

Loading…

From signal to structure

How one recurring pattern in practice became a structured research question.

Example path
Signal
Recurring contradiction in ERP configuration threads
Users resolving symptoms, not the structure beneath them
Observation
Obs 002 — The productivity hypothesis
Not all contradictions should be resolved — some drive the system
Research line
Contradiction lifecycle
Can productive contradictions degrade — and can it be predicted?
Framework implication
Contradiction lifecycle clarified
The line clarified how the pattern behaves across contexts
03

Publications

Findings compressed into transmissible form. Public outputs derived from practice and research — each with a DOI, each citable.

Loading…
Structural class
Series D
When Correction Fails
Extended domain
Series X
Production Failures, Broadly
For delivery organizations
AI in Delivery Systems

How AI shifts delivery from deterministic to stochastic — and what that means for estimation, behavioral contracts, and comprehension at scale.