Decision Engine™ — DecisionForge™ - Decision Application Configuration

Ai Decision Forge™

Ai Decision Forge^™

Loading Decision Application…

Mission Assurance By Design

1. Decision Application

2. Dimension

0 dimensions Choose the type of decision being configured.

Dimension Properties

Validated

Dimension Name

Input Source Type

Data Source / Input

Dimension Role

Data Type

Units

Objective

Importance Weight

0.92

Declared Domain (authoritative)

Server-Derived Admissible Range

Value Authority

Maximum Age (seconds)

Missing-Data Rule

Geometry

Provenance / Source Authority

Observation Rules

Required and freshness checked

Contract range to observed value ± uncertainty

Warning Threshold

Critical Threshold

Behavior Curve

Uncertainty Tolerance

Constraint Protection

Hard constraint / invariant

Decision Explanation

Current Value0.00

Contribution0.000

ConstraintClear

Uncertainty0.00

ProjectionNot derived

AssuranceNotRun

Boundary DriversNone

Generated Curved Decision Surface

The full N-dimensional minimum is calculated first. A deterministic curved projection is then drawn through the dimension probe points and the projected N-dimensional optimum at the selected instant. Lower height means lower objective cost.

Instant of Time

“This curvature was generated by the dimensions, weights, equations, relationships, constraints, and objectives of the decision model.”

Full N-Dimensional Decision Objective

J(x,t) = [Σᵢ wᵢCᵢ(xᵢ,t) + Σ(i,j) rᵢⱼIᵢⱼ(xᵢ,xⱼ,t) + Σᵢ λᵢPᵢ(xᵢ,t) + ρ‖x − x₀‖²] / Σᵢwᵢ

x*(t) = arg min_{x ∈ F(t)} J(x,t)

Displayed Curved 2D Projection

qᵢ(t) = J(x₁*, …, xᵢ(t), …, xₙ*, t)

S(u,v | t) = Σᵢ αᵢ(u,v)qᵢ(t) / Σᵢ αᵢ(u,v), αᵢ = βᵢ/[dᵢ²(u,v)]^1.18, βᵢ = 0.25 + 0.75(wᵢ/w_max)

Calculated Model Values

One additional issue is important: with many decision dimensions, the actual surface is N-dimensional. The 3D image can only display a two-dimensional slice or projection of that larger surface at a selected instant in time. The true minimum must be calculated in the full N-dimensional decision space, while the displayed curved surface explains a selected portion of that result.

Permitted Action Ranking

Only admissible actions are ranked. Prohibited actions receive a score of zero.

Current Decision

Audit / History

Dimension Relationships

Define explicit interactions among dimensions. No hidden learned relationship is introduced.

Source	Operator	Target	Strength	Enabled

Relationship Effect

Enabled interactions alter the deterministic score before actions are ranked.

S(a) = Σ wiPi(xi) + Σ rij

No hidden weights or trained model parameters.

Interaction Summary

Current active relationships in this mission configuration.

Protected Invariants

Safety, authority, protected-zone and abort boundaries cannot be changed by optimization.

Authority Policy

Configure what happens when any protected constraint is violated.

Violation Response

Authorization Mode

Operator Authority

Preserve operator hold and abort authority at all times

Theorem-Guided Certification

Server-derived ranges remain separate from the declared authoring domains. Missing knowledge is never fabricated.

Run Derive Ranges to generate a certificate.

Authorization Gate

Requested values and drone actions are authorized, replaced, held, or blocked from the same certified data.

No authorization result yet.

Derived Dimension Projections

Disjoint intervals are preserved; an inadmissible gap is never represented as permitted.

Dimension	Admissible Set	Preferred	Status	Boundary Source

Solver Evidence Registry

Certificate Integrity and Missing Information