Construction
❯
Building automation
❯Ontologies for smart buildings
Ontologies for smart buildings
For:
Those that can affect the AI system: since it is under the supervision of a
university, the data exchange with the building is controlled by the networking
team of the university and the person in charge of the security. A university
network is not so open! It is not like the Internet that individuals can publicly
access. A group of persons in charge of the GDPR would also be deployed during
the use case.Goal:
OtherScope of use case
Renovation of a building, improvement of the quality of life of the residents
(limited to data issues in the building), audience: citizens, public and private
actors, companies involved in the ICT system managing the building. The scope
is not limited to the building management system (BMS). We would like to open
it to data produced by residents, coupled with data coming from the BMS.
Description
Seminal and technical papers introducing the vocabulary,
definitions, and concepts of smart buildings are References
[205], [206], [207] and [208]. The common view and shared
definition among the community is that a smart building is a
construction with an appropriate design and technological
support to maximize its functionalities and comfort for its
occupants with the compromise to reduce their operational
costs, and extend the life of the physical structure [204].
In Reference [205], the authors presented an initial guide to
understand the layers, taxonomy of services and best
practices for the development of smart buildings. Open
standards are claimed in order to increase interoperability
between layers and services.
In Reference [206], the authors explained variations
between different notions. The findings of the paper allow
the clarification and definition of the border between the
intelligent building and the (more advanced) smart building.
The upper bound of the smart building is defined by (the
future development of) the predictive building. To simplify a
little, from a system point of view, we may think of an
intelligent building as a building reacting to some events
whereas smart buildings are buildings which integrate and
account for intelligence, enterprise, control, and materials
and construction as an entire building system, with
adaptability, not reactivity, at the core, in order to meet the
drivers for building progression: energy and efficiency,
longevity, and comfort and satisfaction.
The INTEL online document [207] is oriented towards the
Internet of things and building management system (BMS).
Analogous to a supervisory control and data acquisition
(SCADA) system used in manufacturing, a building
management system monitors and controls various building
systems, such as heating, ventilation, air conditioning
(HVAC), and lighting with additional and often separate
systems to control elevators, fire, safety, security, and access
controls. We explain later on that our work, at the system
level, is not about BMS, which we consider to be unable to
learn using the data it is managing.
The technical document [208] gives more details about BMS,
direct digital control (DDC), the building automation system
(BAS), and Facility Master System Integrator (FMSI), all of
which are defined according to a system point of view. The
system we propose is more like an operating system for the
building or like an orchestrator of machine learning tasks or
computing tasks and it does not look like any of these
systems.
Finally, the residential buildings system project, from the
Berkeley Lab [197] is also a good source of papers, from
1978 until today, related to smart buildings with a special
focus on the movement of air and associated penalties
involving distribution of pollutants, energy and fresh air.
The ISO process or technology regulations related to Smart
Buildings are ISO 16484-2:2004 [198], ISO 16484-
6:2009[199], ISO 16484-5:2012 [200], and ISO 16484-
3:2005 [201]. They are not related to AI nor to data produced
by residents.
The objective of the use case is to study existing (open) data,
and to build new tools to collect data produced in a building
in order to classify them in ontologies. To be short, an
ontology is knowledge as a set of concepts. The idea behind
the standardization, here, is to put some order in the brute
data and to extract general knowledge. There is a lack of
inclination, in the smart building field, to structure the data
(all types of data) in order to infer and based decisions or
reactions on general knowledge instead of scattered facts.
We are also guessing here that a collective
intelligence/knowledge helps a lot toward good decisions
for people living in buildings.
Sensor Network - IOT
AI: Perceive