HAL Id: hal-00973510
https://hal.archives-ouvertes.fr/hal-00973510
Submitted on 4 Apr 2014
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci-
entic research documents, whether they are pub-
lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diusion de documents
scientiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés.
Toward Web Enhanced Building Automation Systems
Gérôme Bovet, Antonio Ridi, Jean Hennebert
To cite this version:
Gérôme Bovet, Antonio Ridi, Jean Hennebert. Toward Web Enhanced Building Automation Systems.
Big Data and Internet of Things: A Roadmap for Smart Environments, Springer, pp.259, 2014, Studies
in Computational Intelligence. �hal-00973510�
Toward Web Enhanced Building Automation
Systems
G
´
er
ˆ
ome Bovet and Antonio Ridi and Jean Hennebert
Abstract The emerging concept of Smart Building relies on an intensive use of sen-
sors and actuators and therefore appears, at first glance, to be a domain of predilec-
tion for the IoT. However, technology providers of building automation systems
have been functioning, for a long time, with dedicated networks, communication
protocols and APIs. Eventually, a mix of different technologies can even be present
in a given building. IoT principles are now appearing in buildings as a way to sim-
plify and standardise application development. Nevertheless, many issues remain
due to this heterogeneity between existing installations and native IP devices that
induces complexity and maintenance efforts of building management systems.
A key success factor for the IoT adoption in Smart Buildings is to provide a
loosely-coupled Web protocol stack allowing interoperation between all devices
present in a building. We review in this Chapter different strategies that are going
in this direction. More specifically, we emphasise on several aspects issued from
pervasive and ubiquitous computing like service discovery.
Finally, making the assumption of seamless access to sensor data through IoT
paradigms, we provide an overview of some of the most exciting enabling applica-
tions that rely on intelligent data analysis and machine learning for energy saving in
buildings.
G
´
er
ˆ
ome Bovet (1)(2), Antonio Ridi (2)(3), Jean Hennebert (2)(3)
(1) Telecom ParisTech, 46 rue Barrault, 75013 Paris, France
(2) University of Applied Sciences Western Switzerland, Bd de P
´
erolles 80, 1700 Fribourg,
Switzerland
(3) University of Fribourg, Bd de P
´
erolles 90, 1700 Fribourg, Switzerland
1
2 G
´
er
ˆ
ome Bovet and Antonio Ridi and Jean Hennebert
1 Introduction
In the last decade, we have become more and more concerned by the environmen-
tal dimension resulting from our behaviours. Further than the ecological trend, the
interests are also economics-centred due to the raising cost of the energy. Repre-
senting 20% to 40% of the global energy bill in Europe and USA, buildings are
a major source of energy consumption, actually more important than industry and
transportation [27]. In a building, half of the energy consumption comes from the
Heating, Ventilation and Air Conditioning systems (HVAC), followed by lighting
and other electrically operated devices. In offices, HVAC, lighting and electrical
appliances together reach about 85% of the total energy consumption.
For these reasons, the reduction of building energy consumption has become
an important objective. Many works are currently undertaken towards renovation
and improvement of building insulation. The other facet is to leverage on energy
efficiency that involves better usage of HVAC equipment taking into account local
production and storage capacity as well as temporal occupation of the rooms. In
simplified terms, the aim is to optimize the ratio between the energy savings and
user comfort.
This objective requires a clear move from state-of-the-art conventional building
automation systems to advanced information systems leveraging on (1) a variety of
interconnected sensors and actuators, (2) a unified management of heating, lighting,
local energy production or storage and (3) data modelling capacities to model room
usage and predict user comfort perception.
Most automated buildings are currently working with dedicated building net-
works like KNX [3] or EnOcean [2]. These networks are specifically conceived for
the purpose of building automation, including all layers of the OSI model start-
ing from the physical to the application one. In such settings, a central Building
Management System (BMS) is typically connected to the network and manages the
equipments by implementing the operation rules of the building.
In many buildings, we observe the coexistence of different network technologies,
often caused by the installation of new equipments answering specific physical con-
straints, for example wiring or power supply. The protocols are often relying on
proprietary layers and this heterogeneity actually leads to two situations. In the first
one, several BMS are coexisting and share the management of independent equip-
ments, making difficult any global optimisation. In the second one, a unique but
more complex and costly BMS is used where bridges to the different protocols are
integrated. Without prejudging on the strategies of technology providers, we can
reasonably converge to the fact that there is a lack of standardisation in building
automation systems, at the network level and, probably more importantly at the ap-
plication level. BMS could largely benefit of a common protocol stack compatible
with any device.
Thanks to the miniaturization of electronics and increasing computing power,
devices offering native IP connectivity are appearing. Sensor networks based on
6LoWPAN or Wi-Fi are nowadays in competition with classical building networks.
Everyday objects are now able to connect to IPv4 and IPv6 networks, offering new
Toward Web Enhanced Building Automation Systems 3
functionalities like machine-to-machine communications. This has led to the emer-
gence of Internet-of-Things paradigms (IoT), a research field trying to find answers
in how to connect objects to the Internet from the network point of view, i.e. cover-
ing the first four layers of the OSI model.
Going up to the application layer, the heterogeneity problem is even worse as
there are currently no strong standards defining the semantic of the building re-
sources, i.e. an expression of device and service capabilities. The paradigms of the
Web-of-Things (WoT), which can be viewed as the natural extension on top of the
Internet-of-Things, are here proposing to rely on web application standards. Ar-
guably, these standards are more like a set of best practices than real standards. In
the vision of IoT and WoT, any device embeds a Web server offering lightweight
Web services for interaction. The Application Programming Interfaces (APIs) of
the services often rely on RESTful principles, providing natural ways to embed the
semantics in the communication protocol.
In this Chapter we will present the actual state-of-the-art in the field of IoT in
smart buildings, putting into evidence remaining ongoing challenges. Some propo-
sitions overcoming open problematic will be discussed, especially regarding the
integration of existing building automation systems in the core IoT and their respec-
tive discovery. Finally, making the assumption of a building where IoT paradigms
are fully integrated, we will provide an overview of some enabling applications that
rely on data analysis and self-learning algorithms for energy saving in buildings.
2 Integrating Building Automation Systems in the IoT
Many new or renovated buildings are nowadays equipped with automation net-
works. We can here mention office buildings, factories and even private households.
The relative high investment costs have an impact on the payback period which is
rather high, often around ten years. A sudden change of technology is therefore
not conceivable. We envision here the IoT as adapting itself to existing installations
and thus encouraging a smooth transition until building automation systems na-
tively support it. Meanwhile, a mix of different technologies will probably coexist
in buildings.
In this section, after reviewing existing building automation systems and tech-
nologies, we propose a Web-oriented protocol stack able to solve the heterogeneity
problem between sub-systems. Concrete application scenarios will serve as basis of
discussion.
2.1 Related Work
Historically, buildings are equipped with networks especially designed for automa-
tion purposes and offering services tailored to buildings. We can here mention sev-
4 G
´
er
ˆ
ome Bovet and Antonio Ridi and Jean Hennebert
eral technologies like BACnet, LonWorks, KNX and EnOcean. The physical medi-
ums are typically not restricted to certain types, like KNX can support twisted pair,
RF, power line and even Ethernet. Because of the custom protocols used for the
transport and networks layers, it is not conceivable to shift the application layer to
a more common network like IPv6. Some works have proposed to rely on multi-
protocol devices [13]. The drawback of this approach resides in the integration cost
of such devices that are, anyway, quite non-existent on the market. Another solu-
tion consists of providing gateways. As illustrated in Fig. 1, gateways can operate
according to two modes, a N-to-N protocol mapping or a N-to-1* approach map-
ping all sub-systems to only one central protocol. The 1* refers to a new protocol
stack suited for IoT interactions. In the N-to-N case, gateways between BAS trans-
late telegrams from the originating network to its destination. Those gateways have
knowledge about each protocols composing the stacks of each network. Although
this approach solves the heterogeneity across networks, it induces some limitations.
First, it is possible that not all capabilities of a BAS can be mapped to another one,
thus restricting functionalities. Secondly, this approach requires
n∗(n−1)
2
mappings
between BAS, representing a considerable effort.
KNX
BACnet LonWorks
EnOcean
KNX
BACnet LonWorks
EnOcean
1*
Fig. 1 N-to-N (left) and N-to-1* (right) approaches for protocol mapping in buildings.
In contrast to the N-to-N approach, the N-to-1* one considerably simplifies the
number of gateways needed to n by introducing a common technology. The key
challenge resides in the 1* technology where no standard is currently defined. Its
components have still to be identified, according to Internet-of-Things constraints.
A remaining decision has to be taken when integrating BAS into the IoT regarding
the gateway position in the network. There are two extremes, either centralizing the
services on a single node, or migrating them as close as possible to field devices.
Centralizing the access at a backbone server brings some advantages in terms of
maintenance, even if scalability problems may arise. Putting the services at the field
level requires devices with more computational power but allows a direct native
interaction between sensors and actuators with Web services over IP. More specifi-
cally, we can describe four different integrations styles, as illustrated in Fig. 2.
