Lesson 1 of 5 · 12 min
Twin state
Define the state a twin carries and how observations update it.
A twin is a model instance tied to a real or synthetic entity, updated as observations arrive. Its state might be a machine’s health, a customer’s propensity or a patient’s length of stay.
State estimation, for example with a Kalman filter, blends the model’s prediction with noisy sensors.
Key points
- Twins are updated by observation
- State estimation balances model and sensor
Knowledge check
Sign in to mark this lesson complete and resume where you left off.