Meaning
Reconstruction accuracy is measured against the ability of computational techniques to estimate internal system variables that are not directly observable by using the history of available inputs and outputs. In the management of complex organizations, state reconstruction provides a way to infer the health of a project from secondary metrics like velocity and bug counts. The process uses a cardinal vector to represent the current condition of the system in its phase space.
It reaches its limit when the observability of the system drops below the threshold required for a unique solution.
System Observation
Data collection is the foundation upon which the entire inference process is built. When a founder employs state reconstruction, the focus is on capturing the right set of external signals to build an accurate internal picture. Even a small number of well chosen sensors can provide enough information to recover the full state of a simple system.
Observation must be continuous to account for the dynamic nature of the business.
Inference Logic
Mathematical algorithms translate the stream of raw data into a description of the hidden variables. Effective state reconstruction relies on a model that correctly relates the observed outputs to the internal conditions. If the model is flawed, the reconstructed state will lead to incorrect conclusions about the system performance.
Refinement of the logic happens through comparison with known outcomes.
Knowledge Gain
Insight into the true status of an operation allows for more precise interventions when things go wrong. Achieving a successful state reconstruction reduces the uncertainty that often plagues high level decision making. It allows the person in the leadership role to see beyond the surface level reports and understand the underlying mechanics of the organization.
Clarity is the ultimate goal of the reconstruction effort. This technique enables the management of variables that cannot be directly touched.
