Devices, radio technologies and dashboards often come to mind when we think of IoT. However, in a working system, they are only links. The real question is whether the data travels all the way from the physical phenomenon to the appropriate business or operational decision.
The measurement goal comes before the sensor
First, it is necessary to clarify what decision we want to make better or faster. Different accuracy and sampling is required to investigate a comfort complaint, to recognize an energy consumption discrepancy or to monitor the status of an equipment. If the decision is not named, it is easy to create a lot of data with little usable information.
The device must work in a production environment
Measurement in the laboratory and long-term field operation are two separate tasks. Power supply, enclosure, mounting, cabling, radio coverage and subsequent accessibility all matter. An endpoint must not only report a measurement value: we must know when it last logged in, whether its power supply is adequate, and whether we can trust the data.
We choose communication for the location
There is no optimal technology for every project. A wired connection can be stable, but it is expensive to build. Wi‑Fi is readily available, but may be limited by coverage and institutional rules. Radio solutions with low data rates can cover longer distances, but must be designed with different message sizes and response times. A good system buffers, resends and preserves the original timestamp when the connection is lost.
The raw data is not yet business information
The incoming value needs a context: which device, measurement point, room or site, which unit and time does it belong to? Data quality management can also mean range control, detection of outages, duplication filtering and source traceability.
Task from event
A persistent temperature difference should not be just a red dot on a screen. Connect to the affected location and device, get severity, responsible and deadline, and then have a track of investigation and closure.
Not all cases require automatic intervention. The degree of automation should be chosen based on risk, reversibility and operational responsibility.
The life cycle begins after installation
The device fleet changes: the configuration changes, the software is updated, a battery dies or the network changes. Device registration, authorization management, diagnostics and versioned updates are therefore part of the basic system.
Decision checklist
- Did we name what decision the measurement supports?
- Do we know the required accuracy and response time?
- Does the device match the real deployment environment?
- Do we deal with outages and data quality errors?
- Is the data clearly linked to device, location and time?
- Have we determined who and how will react to the event?
The whole chain matters
In the OrigSmart approach, the field device, the data connection, the platform and the business process are part of a system. They must maintain the meaning and reliability of the data at all points.
Let's talk about the measurement task