Impact first, data last.
SMILE — Sustainable Methodology for Impact Lifecycle Enablement — is a design-science methodology that turns any project into a continuously learning digital twin. Its founding move is a simple inversion: the Outcome-to-Data chain (Outcome → Action → Insight → Information → Data) designs downward from intended outcomes to the observations actually needed. Data without an outcome relationship is a candidate for elimination.
SMILE — connecting outcomes, people and AI through reality.
Inspired by NASA JPL's Extreme Collaboration practice — humans, agents and experts in one loop, around a shared, live model of reality, instead of documents thrown over a wall.
Six concentric capabilities — each one keeps revalidating the ones before it.
Every phase is worked from four perspectives: From People · From Systems · From Planet · From AI.
What changed since v5.1.
"Interoperable" (v5.1) becomes "Impact" (v6.x) Lifecycle Enablement.
Governed reality forks with a Reality Consensus Protocol for reconciling them.
Canonical metamodel aligned to PROV-O, SOSA/SSN, OWL-Time and GeoSPARQL.
Two claim states, Unobserved and Believed: an expected observation that does not arrive is an event, never a null.
Hazard-class outcomes carry time-to-impact and time-to-inform per affected actor; closure is a gate condition before authorisation.
A fabric that holds a hazard observation owes a warning to persons within its referent, enrolled or not, with discharge recorded in the lineage.
Freshness Envelope, Partition and Reconciliation Protocol, Degraded-Mode Delegation and a Portable Sovereign Twin Capsule for disconnected operation.
Testable research propositions, not just descriptive claims — the paper can be wrong, on the record.
153 checked against publisher records at v6.4.2; eight added in v6.5.0 resolved against Crossref, the RFC Editor, ISO and NIST on 3 September 2026.
Where the phase logic has been applied.
The current corpus applies the phase logic across eight domains. The record is honest that this demonstrates transferability across contexts — not yet comparative superiority over alternative methods.
Appendix I documents an unmodified 24 Jan 2022 WINNIIO building-data-collector screenshot — persistent GUID, spatial anchor, live CO₂/temperature/humidity/occupancy with provenance — showing the metamodel's core pattern running in production four years before the spec.
Five maturity levels.
Later-phase advancement is gated by integrity in earlier phases — a draft agent while immature in Reality Emulation boundaries shouldn't be described as mature Continuous Intelligence.
Where are you on this ladder?
Your Twin Score on /readiness is this maturity model, operationalised — depth across phases, not isolated AI.
Citation
Waern, N. (2026). SMILE — Sustainable Methodology for Impact Lifecycle Enablement (v6.5.1): Operational Grammar for Auditable, Shared and Forkable Reality. Zenodo. https://doi.org/10.5281/zenodo.20175405
Waern, N. (2026). SMILE v5.1: The Universal Methodology That Turns Any Project Into a Continuously Learning Digital Twin. Zenodo. https://doi.org/10.5281/zenodo.21268264
ORCID: 0009-0001-4011-8201 · WINNIIO AB · CC-BY-4.0
Nobody brings the full stack of digital twin competence — organizational, global, change management, technology strategy, and a NASA JPL-derived method — the way we do. We bring the best people together for a company faster than anyone else, and we implement it with them.
Every quarter without a twin strategy is a quarter of decisions made without a rehearsal.
Scoping workshop from €4,500, credited in full against the project. See how scoping works