When machines learn to communicate: the role of ontologies in the interoperability
In a previous blog post, we talked about the importance of interoperability and how it allows different systems to communicate with each other without barriers. We used the metaphor of the digital Tower of Babel to explain the challenges that arise when multiple technologies, devices and platforms try to share information without common language. In this context, one of the pillars facilitating semantic interoperability is the use of ontologies.
But what is an ontology and why is it so relevant for the digital and energy efficiency world? Let´s explain it in a simple way.
Machine language: how we understand?
To understand what an ontology is, let´s think about how human beings communicate. We do not all speak the same language, and each language has its own grammatical structure, sounds and written symbols. Even within a language, there are dialects and regional variations that can make communication more complex.
Machines and digital systems face a similar problem. Each manufacturer of sensors, devices or software may use its own “language” to represent data. One building´s air conditioning system may report the temperature in degrees Celsius, while another reports it in Fahrenheit. Some devices may call a value “room temperature”, while others simply label it “temp” o “T”. If these systems do not have a common dictionary, communication between them will be difficult or even impossible. This is where ontologies come into the picture..
What is an ontology ?
In the field of computer science and AI, an ontology is a structure that defines concepts and the relationships between them within a specific domain. In other words, it is a way of organising information so that different systems understand it in the same way.
Returning to the language analogy, an ontology is like a multilingual dictionary with clear grammatical rules. It not only establishes equivalences between concepts belonging to different languages, but also establishes the relationships between them. For example, if an ontology says that “room temperature” and “temp” means the same thing, a system using this ontology will consider both expressions as equivalent. Moreover, an ontology allows inferring new information from the knowledge that is already defined in it. That is, it not only stores data, but can also use it to infer things that weren´t explicitly written down.
To fix the concept of ontology let´s imagine a house, in which we could define:
- Concepts: doors, windows, wall, room, kitchen, bathroom…
- Relationships: a door conects rooms, windows are in walls, a bathroom is a type of room….
With all this described and well formulated, an artificial intelligence could answer questions such as, can a window be on the roof? or can there be more than one door in a house?
Ontologies help machines reason about information, allowing them to understand concepts in a more structured way, and not just as loose data. In fact, ontologies are often used in intelligent search engines, robotics, chatbots, etc.
Ontologies and semantic interoperability
As mentioned in our previous post, interoperability has several dimensions: technical, syntactic, semantil and organisational. In this case, ontologies play a crucial role in semantic interoperability, ensuring that systems understand and interpret information in the same way.
Imagine a platform that manages the energy efficiency of a smart building. It receives data from multiple sensors and systems: lighting, air conditioning, electricity consumption, air quality, etc. If each of these devices uses a different way of representing the information, without an ontology to standardise this data, it would be a chaotic to try to process and analyse it in a unified way.
The use of a pre-established ontology will allow this platform to recognise that “temperature sensor”, “thermometer” and “internal climate” are related, ensuring that the information is processed in a consistent and homogeneous way.
“An ontology is a structure structure that defines concepts and the relationships between them within a specific domain”
Ontologies in everyday life and energy efficiency
Ontologies are not only a exclusive concept on digital world. In our everyday life, without knowing, we use similar structures to organize information. For example:
- In a supermarket, products are organised into sections: fruits, dairy products, meat, bakery etc. This scheme helps us to find what we are looking for quickly.
- In a library, books are classified by genre, author and subject, making them easier to find.
- In the medical field, there are classification systems for diseases and medicines so that health professionals speak the same language.
In the field of energy efficiency, ontologies are essential to develop services that turn buildings into smart buildings capable of self-managing and optimising their consumption. By using a common ontology, different systems can exchange information without misinterpretation, allowing lighting, HVAC and other devices to work together efficiently.
In addition, ontologies allow reasoning (drawing conclusions), which facilitates the development of decision support systems to optimise energy use, reduce waste and improve the operational efficiency of buildings.
There are several projects in which CARTIF analyses and applies standard ontologies to ensure that data from different buildings are understandable and reusable in advanced digital solutions, such as the DEDALUS and DigiBUILD projects. In both projects, the use of ontologies allows the information to be unified, thus facilitating the generation of joint building automation and control strategies and decision making based on real data. Furthermore, the use of ontologies allows the different systems being developed in these projects to “speak the same language”, which means that they can easily exchange information and understandeach other, even if they have been designed by different entities or for different functions.
Through the use of ontologies, we incorporate a new technological enabler that allow us to build a more digital and sustainable future, where information flows without barriers and where buildings are truly intelligent, thus contributing to the decarbonisation and sustainability of the planet.