Everyone is writing paper on usage of ontology in IoT. But is there any practically example?. if there is no issue of horizontally scalability, i think there is no need of ontology in IoT.
Frankly, I'm a skeptic, and that's because I do not think there is anything new in IoT. It is merely the natural progression of connecting things together, which the lingua franca of "networking," meaning Internet Protocols, has greatly facilitated.
For various motives, people, including researchers, find the need to give even just evolutionary innovations, a new name. A perfect example is "the cloud." Cynically, I think one of the reasons is just to create some excitement, which in turn means money. Conferences, symposia, grants.
Connecting sensors to a processing unit, and then that processing unit to actuators, with the human being either out of the loop entirely, or perhaps the human only monitoring the process occasionally, has long existed in automated controls. It's been done with purely mechanical and with pneumatic systems, it's been done with hard-wired analog and digital systems, it's been done with a variety of industrial bus standards (CAN, Modbus, Profibus, ...), so now it can also be done over Internet Protocols. In fact, some of these legacy bus standards are very simply layered on top of IP, encapsulated in UDP/IP or TCP/IP packets.
Do we really need a brand new ontology? I don't think so. But we seem to be getting one anyway. The risk being, we can be missing the existing and well established forest, through all the new trees we've planted. For me, things make a lot more sense when they are viewed in context, when the "ontologies" are tied together, in an evolutionary process.
Frankly, I'm a skeptic, and that's because I do not think there is anything new in IoT. It is merely the natural progression of connecting things together, which the lingua franca of "networking," meaning Internet Protocols, has greatly facilitated.
For various motives, people, including researchers, find the need to give even just evolutionary innovations, a new name. A perfect example is "the cloud." Cynically, I think one of the reasons is just to create some excitement, which in turn means money. Conferences, symposia, grants.
Connecting sensors to a processing unit, and then that processing unit to actuators, with the human being either out of the loop entirely, or perhaps the human only monitoring the process occasionally, has long existed in automated controls. It's been done with purely mechanical and with pneumatic systems, it's been done with hard-wired analog and digital systems, it's been done with a variety of industrial bus standards (CAN, Modbus, Profibus, ...), so now it can also be done over Internet Protocols. In fact, some of these legacy bus standards are very simply layered on top of IP, encapsulated in UDP/IP or TCP/IP packets.
Do we really need a brand new ontology? I don't think so. But we seem to be getting one anyway. The risk being, we can be missing the existing and well established forest, through all the new trees we've planted. For me, things make a lot more sense when they are viewed in context, when the "ontologies" are tied together, in an evolutionary process.
the IoT uses various intelligent system components (sensors, actuators, robots, etc.). These intelligent components require a vocabulary and its meaning. The vocabulary and its meaning is generally referred to as ontology. Each IoT area can have its own ontology - in particular the IoT areas such as the Internet of Robotics (IoRT), the Internet of Agriculture (IoA), the Internet of Health (IoH), Green Iot (GIoT) and Wearable IoT (WIoT ).
Ontology is the basis for Deep Learning, Machine / Robotic Learning, Artificial Intelligence, and especially for Cognitive Computing (CC) in IoRT, IoH, and WIoT.
The following sources of literature point to the importance of ontology for IoT:
General about ontology use in the IoT:
J. Munir: "State - of - the - art of Internet of Things ontologies"