When an organization sets out to use the data from its monitoring network, a common confusion surfaces. It assumes that installing the sensors and analyzing what they measure are part of the same job. They are two distinct disciplines, and understanding the difference clarifies where the gap lies.
The physical layer
Installing and maintaining a monitoring network is a specialized craft. The instrumentation vendor selects the sensors, configures them, calibrates them, and maintains them. The systems integrator connects those devices and carries the signal to the system where the data becomes available.
It is field work, with competencies proper to instrumentation engineering and systems integration. Its output is valuable and necessary: without this layer, there is no data.
The analysis layer
Once the data is available, a job of a different nature begins. Centralizing the time series, cleaning them, correcting erroneous values, applying models to detect patterns and anomalies, anticipating behavior, and translating the result into indicators and alerts.
This layer demands data science applied to time series, a discipline that is not the same as the one that installs the sensor. And it is precisely the layer where data turns into a decision.
The gap between the two layers
The problem appears in the space between one layer and the other. The organization that operates the monitoring concentrates its capacity on the physical operation. The vendor delivers capture and connectivity. Advanced analytics is the focus of neither, and the data stops at the chart.
Recognizing that these are different jobs has a practical consequence. The instrumentation vendor and the data analyst do not compete; they complement each other. One builds and maintains the source of the data; the other turns it into knowledge. The organization that understands this difference knows whom to ask for what.
Sources
- EsolverIntegral SAS