Research

Research: the fourth generation

A 3D model shows structure. A 4D model shows change through time. A fourth-generation twin connects observation, state, cause, intervention and predicted futures — and can explain the connection.

Framework

Four generations of digital twins

Each generation answers a harder question about the physical system. Projects are placed by the hardest question they can actually answer, not by how they look.

GENERATION 1

Digital Model

“What does it look like, and where is everything?”

Faithful 3D structure from authoritative data
Facts and physics are known and fixed
No live state, no time axis

GENERATION 2

Digital Shadow

“Where is the system right now, given what the sensors see?”

Observations flow one way into the model
State is estimated from imperfect measurements, with error bounds
Monitoring, not intervention

GENERATION 3

Digital Twin

“What happens if we run it this way?”

A process simulation runs in step with the system
Scenarios, dispatch decisions and control loops feed back
Savings measured against a baseline twin

GENERATION 4

4D Causal Twin

“What changed, when, why — and how does it propagate into every possible future?”

Immutable observed history, estimated present, predicted futures with uncertainty
Counterfactual futures branch from any moment and share one past
Causality derived from trajectory differences, not scripted

Capability matrix

Which questions can the state machine answer?

Use it as a checklist when evaluating any twin — ours included.

Capability	Gen 1 · Model	Gen 2 · Shadow	Gen 3 · Twin	Gen 4 · Causal twin
3D structure from authoritative data	●	●	●	●
State estimated from observations	—	●	◐	●
Process simulation in step with the system	—	—	●	●
Scenarios and control decisions fed back	—	—	●	●
Savings measured against a baseline twin	—	—	●	●
Observed history immutable; prediction recomputed from “now”	—	—	—	●
Branching counterfactual futures from any moment	—	—	◐	●
Causality derived from the data: what, when, why	—	—	—	●
Uncertainty that widens with the horizon	—	◐	—	●

● present · ◐ partial · — absent

The twin runtime kernel

The generational jump is a small kernel, not a bigger framework

Studying the plant and factory twins side by side showed that generation 4 needs a small, domain-free runtime: a pure step function over a plain state, driven by a piecewise input schedule, cached as a trajectory whose past is immutable and whose future is recomputable, branched by structural sharing, with causality read off trajectory differences. The midstream twin validated it in a third domain.

KERNEL · ~500 LINES

Pure stepper · S × Inputs × dt → S
Input schedule: piecewise inputs by time; change from day D, keep everything before
Trajectory: recompute from t, clone with shared history
Variable registry + divergence detection → events, time-to-effect, explanation chains
Deterministic RNG + invariant checks: same seed, same future

Reports and data

Every claim, with the data it came from

Reports are generated from the project repositories at build time; each states what was computed and what is illustrative.

GEN 4 · ENGINEERING REPORT

Have a system you want to see, question or rehearse?

Tell us about the site, the process and the decision. We will come back with what a twin could answer, which data it needs, and how fast a prototype can be in front of your team.

Talk to us All projects