An observation is a result of the data that the sensors collect and that the Discovery environment filters to the IFS back-end systems. An observation can be unknown, received, waiting for review, not executed (marked as error) or completed. The user can also manually set observations to be ignored. The observations comes in from sensors that send information automatically. But if the observation is not executed correctly or is waiting for manual review the user needs to execute all operations within the observations manually. The user can mark the observation as false if the data or the observation itself is not correct. The user can do this by a right mouse button click in the details form of the observation.
If it’s an unknown observation the user can follow the steps in the process mapping in order to define the observation. In the Observation Flow it is shown that an observation can include static and dynamic fields. Static fields are fixed and always there, they are automatically added by IoT Controller when defining an observation flow. They contain necessary information for the system to handle the observations and know what to do with it. Dynamic fields are added by the user to map and describe the information and readings that come in through the observation, that are unique for that observation code.
An observation has a minimum number of fields that should be passed to the IFS back-end systems. Those are Observation Code and Device ID - typically Occurrence Date Time and Observation Message ID are also included. It's also configured in the Discovery environment where a number of additional fields can contain values of different types.
A new Observation Flow is created after the user has registered the observation code from an unknown observation. New Observation Flows can also be registered manually, before the observations start being received.
A new configured Observation Flow is created after the event is published and a operation and possibly condition is added to the observation flow.
Observations that have gone through all the Basic Data steps and where the user have manually reviewed the observations that did not execute correctly or were waiting for manual review.
Observations that is passed to the back-end system, but later is discovered as false - faulty reading, measurement that was not supposed to be done, etc. - can be marked as a false observation. These can then be reviewed by the user in the IoT Controller for further analysis and action.