Why isn't saref:Meter a subclass of saref:Sensor ?

Class saref:Meter is defined in SAREF4WATR as a subclass of saref:Sensor, which is not an axiom that is part of SAREF Core and could be considered hijacking the concept.

added clarity issue label

I Agree. From a previous discussion we decided to remove this subclass axiom.

True. But I hope someone can provide convincing rationale for having them as siblings in SAREF in the first place.

The convincing rationale is that something that is a meter has the same capabilities that something that is a sensor; therefore, a meter should be a subclass of a sensor.

There is the need for a convincing rationale of not having them as subclasses.

I understand a meter as providing a human-readable range of possible individual sensor measurements over a period of time, which is why meters are usually more expensive than simple sensors. A meter 'contains' or connects to at least one sensor. I'm not sure that sub-classing is the right relationship between Sensor and Meter.

Now that we are adding local restrictions to different classes, I think that the question of whether a meter is a sensor or not is clearer.

In the current definition, a meter must observe at least one "thing". So it is expected to use the saref:observes property.

However, (just focusing on devices) the definition of saref:observes is only applied to sensors (as the definition of saref:controls is only applied to actuators).

I see two alternatives (from which I prefer the first one):

• That saref:Meter is a subclass of saref:Sensor.
• That the restrictions on domains and ranges are removed (which I would try to avoid to increase their understandability).

A decision on this issue is required, since it significantly affects at least SAREF4WATR.

mentioned in issue #38 (closed)

Again, I'm fine with that. if someone from TNO (@bouter ) or AI4People (@dragoni ) also agrees, then we can implement the change

(That saref:Meter is a subclass of saref:Sensor.)

I agree as well about defining saref:Meter as a subclass of saref:Sensor.

@bouter, any comment from TNO's side ?

When creating SAREF, we analyzed many existing standards/protocols with the result that sensors and meters are used as different entities, as we then encoded in their definition. Sensor: A device that detects and responds to events or changes in the physical environment such as light, motion, or temperature changes. Meter: A device built to accurately detect and display a quantity in a form readable by a human being. They can both detect quantities, but the meter is made to measure it and display it to the user, while a sensor can detect a quantity, but usually does not contain a display to show the measurement to the user. Conceptually, a meter is not a sensor, can you say that your energy meter at home is a sensor?

Summary 2025-01-20:

Agree to implement a change in the definition:

Sensor: "A device designed to observe and measure one or more properties or states of one or more features of interest." (current definition unchanged)

Meter: "A device designed to observe and additionally do some computation and/or display one or more properties of one or more features of interest."

closed

moved to saref-core#70 (closed)