Research Platform Enterprise Docs
FSI Behavioural Model · Vol. 1

Behaviour modelled as
physics, not personality.

We borrowed from Fluid–Solid Interaction mechanics to build a formal model of human decision-making. An individual is a solid body with defined structure. A decision is a fluid — it flows, exerts pressure, and reshapes what it contacts. The boundary between them determines everything.

Model FSI-v0.4
Status Active Research
Updated Q1 2026
Beta Q2 2026
The FSI Framework

The physics of
human decisions.

In classical Fluid–Solid Interaction, a rigid body is immersed in a flowing medium. The fluid exerts force on the solid; the solid deforms the fluid's path. Neither remains unchanged by contact.

We apply this directly to behaviour. The individual is the solid — a structured entity with measurable density, rigidity, and surface properties. The decision is the fluid — dynamic, pressurised, seeking the path of least resistance through the boundary layer.

The boundary condition governs permeability: what can pass through, what is blocked, and at what threshold. The system is the containing medium — the organisation, the team, the environment — that must absorb and equalise whatever pressure the interaction generates.

FSI Analogy Map
Solid body The Individual — structured, stable, property-bearing
Fluid medium The Decision — dynamic, pressurised, path-seeking
Boundary layer Constraints — permeability, resistance, surface tension
Containing medium System — pressure equalisation, cascade, diffusion
Object A · The Solid
The Individual
A structured agent with defined properties: cognitive density, agency, risk surface, and motivational rigidity. Like a solid, the individual resists deformation — but absorbs stress and changes state under sustained pressure. Profile updates are the model's equivalent of material fatigue.
Object B · The Fluid
The Decision
A pending event that exerts continuous pressure on the individual. Decisions have viscosity (complexity), flow rate (urgency), and temperature (emotional charge). They seek resolution — the path of least resistance through the boundary. Once resolved, they dissipate from the model.
Boundary Condition · Interface Layer
Permeability & Resistance
The surface properties that govern whether a decision can reach and act on the individual. Access rights, social capital, competing obligations, and resource availability act as permeability coefficients. A high-resistance boundary can hold pressure — until it cannot.
System · Containing Medium
The Ecosystem
The ambient environment — team, organisation, infrastructure — that contains and is affected by the interaction. When a decision resolves, displaced pressure propagates outward as a cascade wave. The system must absorb, dissipate, or redirect this energy. Some systems fracture.
Live Scenario · Security

An insider threat
runs the model.

Marcus is a senior security analyst with privileged access. A financial stressor has elevated his risk surface. A decision approaches: exfiltrate proprietary client data to a competitor. Step through each phase of the FSI model.

FSI System State
Analysis Log
Methodology
01 —

Solid profiling

The individual model is never static. Like a material under repeated stress, properties shift: agency hardens or softens, risk tolerance expands under financial pressure, loyalty weakens under prolonged threat of displacement.

  • Bayesian property updates per event
  • Material fatigue modelling over time
  • Phase-transition detection at threshold
02 —

Fluid pressure scoring

Each pending decision is scored on viscosity, urgency, and emotional temperature. High-temperature, low-viscosity decisions move fast and penetrate weak boundaries easily. We compute the pressure differential at each boundary layer.

  • Viscosity: decision complexity index
  • Temperature: emotional charge coefficient
  • Pressure: boundary differential score
03 —

Cascade propagation

Resolution generates a pressure wave. We model its propagation through the containing system — who absorbs it, who amplifies it, and where fracture points exist. Second and third-order effects are the hardest to detect and the most consequential.

  • Agent graph propagation model
  • Fracture point identification
  • Dampening and amplification factors