Predicting material properties
before fabrication.
Predictive Analysis as a Service is the operational layer of the Grunuss architecture. Material behaviour is modelled at electronic scale, validated against measurement, and delivered as institutional reports that conform to the whitepaper template documented under Methodology.
§ 01 / Premise
GS-2026 / SECT_01
Prediction earns the right to direct fabrication.
Most industrial materials decisions are made empirically. A candidate is fabricated, characterised, and either accepted or rejected. The cycle is expensive and slow. Where simulation is used at all, it tends to confirm decisions already made on other grounds.
PAaaS inverts this. Material properties are predicted at electronic scale before fabrication. Predictions carry documented assumptions, declared uncertainty bounds, and explicit validation conditions. When the artefact is later measured, the gap between predicted and observed is published. Confidence is earned, not assumed.
§ 02 / Capabilities
GS-2026 / SECT_02
Three capabilities operating as one engine.
The same simulation engine answers three families of question. Each is bounded by validated method, declared assumptions, and conformance to the institutional whitepaper template.
- C.01
Properties
First-principles prediction of bulk, surface, and interfacial properties. Electronic structure, thermal response, mechanical envelope, electrochemical behaviour.
01 / 3 - C.02
Envelopes
Operational windows under realistic conditions. Stability across thermal, mechanical, and coulombic regimes. Failure thresholds declared in advance.
02 / 3 - C.03
Lifecycle
Degradation pathway prediction and longevity bounds. Property evolution under operating cycles. Validation against measurement.
03 / 3
§ 03 / First live capability
GS-2026 / SECT_03
First live: Battery Life Cycle Prediction.
The three capabilities above, deployed against a single industrial question.
The first capability brought into operational service is Battery Life Cycle Prediction. The same engine that predicts properties, envelopes, and lifecycle for arbitrary engineered materials is now bounded to one application: predicting how a battery system degrades under realistic operating envelopes, and how long before specified performance thresholds are crossed.
Properties prediction (C.01) supplies the underlying material behaviour. Envelopes (C.02) define the operational regime. Lifecycle (C.03) is the integrated output. The work is conducted under the Sim-B program (Simulation – Battery Systems) documented under Research § 04. Future capabilities — superhydride stability, generation-side materials, additive process windows — extend the same engine to other applications.
| Code | Application | Status | Programs |
|---|---|---|---|
| F.01 | Battery Life Cycle Prediction | Live | Sim-B (Simulation – Battery Systems), QMS-B (Quantum Material Simulation – Battery) |
| F.02 | Superhydride Stability | Forming | RTS-W (Room-Temperature Superconductor Workstream) |
| F.03 | Generation-side Materials | Forming | QMS-G (Quantum Material Simulation – Generation) |
| F.04 | Additive Process Windows | Forming | Process-side workstream (not yet named) |
§ 04 / Why PAaaS
GS-2026 / SECT_04
What this platform is, and what it is not.
What PAaaS is.
- W.01
A validated prediction engine
Outputs carry parameter listings, model versions, convergence states, uncertainty bounds, and explicit success/failure criteria. The validation discipline is documented on Methodology § 05.
- W.02
An institutional research surface
Releases conform to the whitepaper template documented on Methodology § 06. Every result is auditable. Sections that cannot be filled honestly are not filled.
- W.03
A long-horizon engagement
Engagements run on cycle, not on quarter. The platform refines through deployment feedback — observed-vs-predicted closure improves the engine for every subsequent partner.
What PAaaS is not.
- N.01
Not a forecast tool
PAaaS does not project market outcomes, demand curves, or strategic scenarios. It models physical behaviour under declared assumptions.
- N.02
Not a black-box service
Every result is delivered with its method, its assumptions, and its limitations. Surrogate-derived outputs are explicitly flagged.
- N.03
Not an unconditional commitment
Engagements are governed by the alignment filter documented on Partnerships § 03. Prospective work is reviewed for ethical compatibility, technical integrity, and strategic coherence before commencement.
§ 05 / Surfaces
GS-2026 / SECT_05
Access surfaces.
Four interfaces, one platform. Each surface exposes the same underlying engine with the same reproducibility guarantees.
- SF.01
Console
Authenticated web UI for program configuration, run inspection, and artefact retrieval.
- SF.02
API
Versioned programmatic interface for solver invocation and artefact access.
- SF.03
Repository
Reproducibility artefacts: versioned inputs, solver builds, seeds, and outputs.
- SF.04
Reports
Structured technical notes with first-principles grounding and uncertainty bounds.
§ 06 / Access
GS-2026 / SECT_06
Access and engagement.
Access is granted to qualified industrial and institutional partners after alignment review. Engagement follows the structured intake documented below.
A.01
Initial enquiry
Submit an engagement profile to research@grunuss.com. Include problem statement, materials and operating regime, validation data available (or absence thereof), and target timeline.
A.02
Alignment review
The enquiry is reviewed under the alignment filter (Partnerships § 03). Outcome typically issued within ten working days. Where alignment is uncertain, a clarifying conversation precedes the decision.
A.03
Workspace provisioning
Confirmed engagements are provisioned on the platform (Workspace). Access carries documented onboarding, methodology orientation, and a defined first cycle.
Closing
PAaaS is the institutional platform, made operable.
The architecture documented across Philosophy, Architecture, Research, and Methodology converges here — the platform that earns the right to make claims about industrial materials.