121
3.
Smart home environment
3.3.3 Integration and interaction
A major problem with smart home design is the integration and interaction among
heterogeneous subsystems, which may probably not be designed to interact with
each other. Assistive technologies are very heterogeneous when attending needs
due to individual and temporal variations. Moreover, devices are usually designed
by different manufacturers using different technologies for heterogeneous
applications. The Design-for-All concept considers the lack of simplification usually
made when considering a standard user. At the same time, this lack of
standardization and individual diversity and variability increases heterogeneity in
subsystem development, both in terms of applications and services, in a kind of
vicious circle. The result may be called "islands of functionality": solutions adapted
to specific users in particular environments.
A smart home should be able to support the interaction of heterogeneous devices,
networks, services and applications. First, there is a need to interact at the
internetworking level. At this level, the necessary mobility of the user implies that
interconnectivity cannot be guaranteed at all times so communications should be
asynchronous. Clients asking for a service and devices offering it may not be
connected at the same time. Therefore, the communication paradigm should be
connectionless (vs. connection oriented), well suited for intermittent connections.
In smart homes, if a backbone fixed infrastructure is available then a Nomadic
system may be better than an ad hoc system: mobile devices connected through
wireless links to a fixed wired network. For instance the backbone network may be
based on the IP protocol, a robust and contrasted solution, which has
demonstrated its success in the interconnection of heterogeneous devices (a good
example is the Internet). Most devices can be connected through this IP network
while secondary, maybe simpler, devices (e.g. sensors) may be connected using
non-IP communications. In this case a gateway is used to interconnect IP and non-
IP sub-networks. This solution permits environmental control in a remote mode via
a web page as well as direct Internet access in home automation through
residential gateways.
Second, interoperability should include dynamic service discovering (periodically or
triggered by determined events), service description (including actions that may be
performed, properties that may be useful), and service control (actions and
modifications of state or attributes of a service in a sub-network from another
device connected to a different sub-network). Services and information from a
given subsystem should be described using common languages and media formats
to be accessible to other subsystems. I n t e raction between context-awa r e
subsystems requires common context representations that are independent of the
122
3.
Smart home environment
applications. But it is not just a simple problem of using a common format. Further
issues are how this context information is interchanged among subsystems, how
services are discovered or offered, and how they are integrated in user interfaces.
A number of available architectures can support these functions: HAVi, Jini, UPnP
(Universal Plug and Play).
This interaction is usually a syntactic intera c t i o n , without considering the
“meaning” of discovered and shared services. However, in the future a higher level
of semantic interactions may be possible allowing services to be pre-selected to
offer adapted assistive services to the right people or to adapt or empower the
functionality of existing applications according to new services. New applications
may become available based on the new services.
A typical example of this interaction could be a sensor network monitoring
physiological parameters (heart rate, blood pressure, sugar level). Some of the
sensors may be body-worn, using a low-rate WPAN like 802.15.4. Others, may be
integrated into the surroundings (chair, bed, building), probably connected through
a backbone network like EIB or a WLAN like 802.11g. One of the body sensors may
act as a bridge to the ambient sensor network, providing interconnectivity at the
network level. But additionally, these devices may interoperate themselves at a
higher level. As an example, in a health monitoring application alarms or drug
doses may be adapted using information from the ambient sensors, for instance
inferring the user activity state (e.g. driving, sleeping, exercising, or talking to
someone).
3.3.4 Wired versus wireless
Wireless technologies have clear advantages and drawbacks when applied to the
smart home environment. Among the advantages, flexibility and easy installation
are clearly important characteristics in this type of networks. Among the
drawbacks, clearly safety and security can’t reach the levels which can be obtained
with wired networks, deterministic response times are not possible and RF
emissions might cause some user concern.
Research investigating spread spectrum techniques in the 2.4 GHz range which
allow a protected transmission to solve the security problem are being conducted
[Fellbaum, 1999, van Berlo, 1999 and Flikkema, 1997]. They are robust against
sinus- and noise-like disturbance sources and wall reflections and they avoid
crosstalk effects between the different RF channels. There is only a serious problem
if a microwave oven is used. This device produces a wide band disturbance signal
which can easily blot out the control and communication signals, even if the oven
is carefully shielded.
123
3.
Smart home environment
However, it is clear that, in many cases the advantages overcome the drawbacks
and wireless network become the most feasible alternative for home automation.
Originally wireless smart home networks were based on protocols specifically
designed for this purpose but currently, due to the huge penetration of computer
and telecom wireless networks, this is not always the case. To be more specific,
three different network families are currently used to support smart homes:
• Traditional RF Home automation networks: These are usually based on
relatively low frequency carriers and modulation techniques are usually quite
basic, thus available bandwidth is usually very small (a few kbits/s or lower).
Examples of these protocols include X10 over RF (at 200MHz) or KNX over
RF (at 868MHz). Many proprietary networks based on RF remote control
frequencies (433MHz) are also widely used
• Wideband RF protocols: These protocols were originally designed for
computer networks and provide relatively high bandwidth (currently up to
hundreds of Mbits/s). They usually operate at 2.4GHz or 5GHz. Currently the
most popular among these type of networks is the WiFi family (IEEE
802.11a/b/g). These networks are very useful for relatively complex devices
but for simple devices, the costs per node and especially power consumption
rule them out
• Generic Low power networks: These networks have been designed very
specifically for mobile device and optimized for low power usage. Bluetooth
is currently the most widely used but its protocols are relatively complex and
its power requirements are not suitable for devices that have to run on a
small battery for years (or get the power somehow from the environment).
Zigbee is a new type of very low power, low complexity network with some
built in localisation capabilities that seems to be very promising for smart
home applications.
Probably in the near future most smart home networks will be based on a mix of
WiFi and low power networks interoperating possibly with some wired segments
as well.
3.3.5 Speech technology
The principles of electronic speech processing as well as applications for persons
with disabilities have already been presented in Chapter 2.2.2. In this section the
focus will be on the use of speech technology in smart homes, especially for elderly
people.
124
3.
Smart home environment
A serious problem of a lot of technology in the home environment is an adequate
control of these systems. Let’s consider a person with some mobility problems,
sitting in a chair in front of a TV set. Usually a remote control is available. A first
idea could be to extend the number of keys of the remote control in order to add
more control functions. However, with more keys the number of malfunctions will
significantly increase and, especially with elderly people, the user acceptance will
dramatically decrease. With a voice controlled system, virtually all control functions
can be executed by voice, which also includes a voice output (with the aid of a
small built-in loudspeaker), a visual component is however indispensable to remind
the user about the functionality he is controlling. The right ‘mixture’ of audio and
visual information, in other words, the ergonomics, has a strong influence on the
acceptance by the user [Hampicke, 2000].
In an Ambient Intelligence environment (see chapter 4) walls and objects of
everyday use have intelligence. In the case of speech processing the walls have
built-in microphones, mostly arranged in arra y s, l o u d s p e a kers and enough
processing power to fulfil all speech processing activities being used for
recognition and synthesis: the walls can listen and speak. Moreover, displays which
are needed, also in a speech dialogue application, can appear everywhere on the
walls. The above scenario, which may sound like science fiction, is quite feasible
today.
While speech recognition tries to detect the content of a spoken utterance, speaker
recognition investigates the identity of the speaking person. The main application
for speaker recognition is access control. It can be the access to houses and/or
rooms or the access to computers or household appliances (including voice-
controlled cooking or washing). The reliability of systems that are on the market
sometimes isn’t sufficient. The reliability of the total system can be improved by
combining speaker recognition with other forms of control, for instance face
recognition by a camera.
3.4 Products and services
Within the field of smart home technology products and services play an important
role in creating benefits for users. In general products and services can be divided
into six categories:
1. Comfort
2. Energy management
125
3.
Smart home environment
3. Multimedia and entertainment
4. Healthcare
5. Security and Safety
6. Communication.
The division is not strict, the different categories overlap. Mostly people don’t
choose just one category but several. The different categories strengthen each
other leading to the fact that overall functionality of several categories combined
is more than the sum of functionality of the independent categories.
It is impossible to give an extensive overview in this chapter, or anywhere else for
that matter, since new products and services are developing at such a rapid pace
whilst by combining solutions new products and services are formed.
The emphasis in this section is on products and services belonging to the
categories healthcare, security and safety and communication, giving some
examples of solutions that fall within the different categories. For a broader
perspective a list compiled by the Ambient Assisted Living Initiative is available
[Steg, Strese, Loroff, Hull & Schmidt, 2006].
3.4.1 Healthcare
On a global level healthcare solutions can be divided in the following categories:
• Active alarm systems: utilize remote emergency systems – usually telephone
based – installed in the home of older persons
• Passive alarm systems: do not require the interaction of the person. For
example, devices include sensors that are able to recognize the danger of a
fire and send an emergency call automatically
• Remote support for care staff: include all kinds of telecommunication-based
activities supporting the work of field staff
• Remote support for family carers: includes all kinds of telecommunication-
based activities supporting family carers
• Advanced services using video telephony: include remote monitoring and
video-based alarm services
• Telemedicine
[European SeniorWatch Observatory and Inventory, 2002].
126
3.
Smart home environment
Alarm systems play a very important role in Smart Home applications. However, the
systems, existing so far, are very often too complicated or not reliable enough. The
key issue is the setting off of alarms. If, for instance, persons have an accident (fall,
injury), a fire breaks out or a person suddenly becomes unconscious, then there is
normally no time or possibility to operate a telephone or even an alarm button on
their wrist or around their neck. Although there were many alarm systems on the
market and several research projects have looked into this particular issue (e.g. the
TIDE projects FASDE and ASHoRED) exist, there is still a large knowledge gap as to
how persons react in a dramatic situation, be it panic or a collapse or simply
because they are confused or have memory problems (forget that they are wearing
an alarm button).
Probably the best solutions to overcome these problems are passive systems.
Passive means an automatic control of vital functions (e.g. pulse, blood pressure,
oxygen saturation), their evaluation and an automatic alarm being set off when the
values of the vital parameters exceed predefined limits. The reliability of the alarm
being set off can be dramatically increased when several different observations and
decisions are combined. Among the measurement of vital parameters as described
above, the person’s activity (leaving and entering rooms, using water, electric light,
TV and radio and many more) can give important additional information.
In order to avoid false alarms the receiver of the alarm (e.g. call centre or relatives)
sends back a signal to the user for instance by telephone or via a message that
appears on the alarm module worn on the wrist or around the neck. In the case of
a false alarm, the user can answer the phone or press the button of the module
indicating that no help is needed thus avoiding unnecessary attendance. If, on the
other hand, the user does not react, it can be assumed that the alarm is serious
[Hampicke, 2004].
One huge problem in healthcare is wandering and wayfinding. There are systems
to detect where a client leaves the house where this would be inappropriate or
dangerous for the client. These systems consist of magnetic contacts or pressure
mats at/near hall door connected to local area (family) paging. The system, of
course, does not restrict egress, but merely alerts a carer that the client has left.
Similarly, for wayfinding (at night) lighting strips and passive infra-red light
switches can assist and reduce the likelihood of a fall or disorientation around the
house at night.
An area that receives an increasing amount of interest is telemonitoring or
personal health monitoring is based on the idea that persons can monitor
themselves in their home using medical devices. Health care monitoring at home
enables continuous measuring of physiological parameters. It is possible to embed
127
3.
Smart home environment
sensors in different places or objects at home (e.g. in the furniture, electrical
appliances), or to make them wearable by integrating them into clothing "Smart
Shirt" or small apparel items such as watches and jewellery. By combining these
wearable sensors with measurement devices embedded in home surroundings,
advanced multiparametric health monitoring may be achieved [Korhonen, Pärkkä
& van Gils, 2003].
Recording of physiological and psychological variables in real-life conditions could
be especially useful in management of chronic disorders or health problems; e.g.
for high blood pressure, diabetes, anorexia nervosa, chronic pain, or severe obesity
[Korhonen, Pärkkä & van Gils, 2003]. Telemonitoring could also be used to provide
feedback about someone’s health in the form of behavioural feedback in order to
prevent diseases.
Obviously telemonitoring has many advantages for both, the patient and the
medical institutions. The patient can stay at home and does not have the
inconvenience, associated with a visit of the doctor or in the hospital, and the
medical institution saves time because there is no need for spending time with
routine work and in the hospital enormous costs for the bed and the care of the
medical staff can be saved. In several telemonitoring applications the data are not
directly transmitted to the medical institutions but to a kind of ‘call centre’, which
performs a first data evaluation.
3.4.2 Security and safety
Top priority for many older people is the feeling of living safely and securely in their
own house. In general, people like to know who is at the central access door of the
flat and at the front door of the own apartment, before opening the door. In many
projects this access control has been facilitated via remote control by phone, on TV
and electronic locks on central access door and own apartment door. In some
projects an intrusion alarm is present. Residents have to enter a code or use a
proximity key to switch off the alarm.
A smoke detector is installed in all projects, as near as possible to or in the kitchen.
In some cases there are even smoke detectors in kitchen, living room and bedroom.
If smoke is detected an alarm signal is given to a call centre automatically. First,
the call-centre operator will speak to the resident via the safety alarm phone if
there is a real fire. The smoke detector can be accompanied by other sensors to
create an even safer environment.
Older people have a more frequent nightly toilet visit than younger people. With
automatic light switching on when the legs are put out of the bed, one can better
orientate and find the way to the bathroom without risks of falling.
128
3.
Smart home environment
In most cases the electric or gas cooker can be switched off via an extra button,
which also switches off the light on the working area. On the other hand the
cooker cannot be used if the light on the working area is not switched on. There
are also examples where there are two switches: one switch to activate the cooker
and one to turn on the light. If the resident leaves the house, the cooker is
automatically switched off. The same is true when the tenant goes to sleep and
uses the button "day/night" above the bed or in case of an alarm.
3.4.3 Communication
One of the most commonly used smart home technologies is internal and external
Intercom systems, providing elderly and disabled people with a method of
communicating with those calling at their door and to other rooms in the home.
Many mobility impaired persons will require an intercom system in order to safely
answer the front door without having to go to it. Going to the door may be (a)
impossible or (b) difficult or slow causing danger transferring to a wheelchair or
missing the caller due to the delay. Using CCTV in addition can enhance personal
security when admitting visitors and may also provide verification that the visitor
has actually left the house if they have been visiting a person in an inner room.
The video signal can be distributed to a TV or any other kind of screen within or
outside the home. It is also possible to send the signal from the front door to a call
centre where a person at the centre can talk to the visitor. This could be helpful
when it is late in the evening and the resident feels insecure when answering the
door bell.
Communication isn’t restricted to ‘normal’ activities in the home environment it
can also be used for remote learning. In order for remote learning to be successful,
the technical infrastructure must be in place. In today’s technology, this means the
student should have access to a wireless network, web cams, a microphone,
keyboard access and full telecommunication infrastructure linking the home to the
outside world. All of these can be set up exactly to the user’s requirements, in a
manner that would be extremely difficult or even impossible in a classroom
environment.
Remote learning carried out in a smart home would need to take place as part of
a designed educational system, and could not be put in place piecemeal. An
assessment of the needs of the student would also need to be carried out to insure
that the student will be able to access all parts of the course he or she has enrolled
for.
129
3.
Smart home environment
Once the correct structure is in place, the potential benefits are impressive; a
student will have access to the expertise needed to complete the course, the
student experience will have an immediacy to it, as the student would not need to
invest time and energy travelling to a class. Their needs would be addressed in a
very individual way. Education delivered to a smart home would overcome
geographical or environmental barriers to such services. It would allow direct
communication with the instructor, and gives access to resources immediately:
educational as well as personal, health and social care. Because a smart home also
gives access to wider society (for example, e - g o v e r n m e n t , shopping and
entertainment), a student would have more time available to him or her, so their
quality of life would improve.
For many, accessing good and reliable information about suitable technology is a
significant barrier to using technology that meets their individual needs. The
Central Remedial Clinic in Ireland, provides a vendor neutral information and
evaluation service. People with disabilities, their families, carers and other service
providers can get up to date information about the technologies currently available
in Ireland and can receive guidance in terms of how such technology may be
installed and set up for use in their own homes. Staff at the Central Remedial Clinic
works to identify the reasons why a person is seeking such technologies and then
devise a solution based on the supports that are available to the person and
contextual and environmental factors such as the location of their home. Smart
home solutions usually comprise recommending a range of technologies that are
to be placed in a person’s home environment and a method of controlling these
that is suited to a person’s individual needs.
3.5 User interaction
The advent of smart homes is part of the overall pattern of convergence that is