The descriptor of actuators is defined solely by the type and model of the device that is controlled. This area is not well covered by any existing ontology and the broad scope of potential devices that could act as actuators in the network means that developing an ontology would best be achieved through a long-term multi-partner effort. In order to both maintain standard behaviour within the OSDL Descriptor Environment and allow for further adoption of a well-recognized ontology when this becomes available we use URLs to define the type of device being controlled (using DBpedia) and the model of device used (from the manufacturer’s website).
|name||Human readable name of the controlled device|
|deviceType||URL to category of device from DBpedia|
|deviceModel||URL to device page on manufacturer website|
|target||URL to either variable affected (e.g. temperature) or the result of the action (e.g. a photograph)|
In some instances it is necessary or beneficial for actuators to make a return response. For example if a GoPro camera is triggered to take a photograph (see GoPro Controller example) it is possible to get the URL of the image as served by the camera over WiFi. This allows the image to be associated with data from sensors, such as environmental conditions.
Actuators can respond with a JSON formatted array of URL=>value or URL=>array values, keyed by the type of response. A photograph from a camera would be keyed by the descriptor URL http://dbpedia.org/resource/Photograph. The use of standard URL schemes for identifying the type of returned content allows for Controller nodes to perform media type specific operations, such as automatically geotagging photographs by adding EXIF metadata or uploading videos to sites such as YouTube or Vimeo.