The 'smart home' is term bandied about, that conveys a number of meanings and implicit assumptions. A similar term 'the intelligent home' conveys a set of meanings that are implicitly inaccurate and therefore, this paper contends, we should rethink our notions of smartness or intelligence in order to present a uniform appropriate name for the technology that we currently use.

What do we mean by a 'smart' or 'intelligent' home?

David Gann et al (1999) suggest that a smart home is about "using the latest information and communication technology to link all the mechanical and digital devices available today - and so create a truly interactive house" (Gann et al, 1999, ix). If this were the case then there are no smart homes as no home effectively manages to connect all the mechanical and digital devices available today, as that would be a simple impossibility as well as requiring a Very large house.

The whole notion of smartness and intelligence is considering the technology from the position of what the technology is like and personifying it, making it smart or intelligent, where as we should be considering people within the context of the name. What does the technology do for the people that use it?

Fair enough, there are a number of users of technology, well in fact everyone uses technology in some form or other, but specific people use 'smart' home technology. Firstly, there are the 'Gadgetheads', the people who will use anything new, just because it is there and a new gadget. These are the home automation enthusiasts who build their own X10 systems. Secondly, there are the 'Glamourheads' of the world who need the technology in order to maintain a quality of life and status (otherwise they would have to open the front door themselves!). They use technology to enhance their standing in the community and to protect their property and privacy, as well as run their baths and heat up their fish and chips. Thirdly, there are the 'Ecophiles', who use this form of technology to save planetary resources by ensuring that the system modifies the environment in their living spaces for optimum efficiency and minimal wastage. These systems have the ventilation and heating systems combined to produce the most efficient ecological benefits.

Finally, there are the people who need the technology to support their lives and lifestyles, who are dependent on the technology for a quality of life. Disabled and older people might benefit from the use of technology in many ways, but their relationship to it is significantly different from the other users. This final group are dependent on the technology. Without it, the quality of their lives would be seriously diminished. They might use the technology to support them in their homes in a number of different ways such as to assist in reminding of events and activities, or alerting other people if they are unwell or have fallen. The dynamics of the 'smart' home for these people range from simple small systems of a few devices configured together through to elaborate systems monitoring bodily activity and sensors throughout the house.

I see a considerable difference in the relationship that this group have with technology from the other groups. Older and disabled people are dependent on the system. It must work all the time, every time, in the way it is supposed to work, otherwise the people using it will stop using it or could be in serious problems. For these people the technology is not smart, not intelligent, it is part of their life, just as a telephone is to most of us. We do not consider the phone to be smart or the refrigerator to be intelligent ('Smart fridges' excepted) we just consider them as parts of every day life.

The term 'smart' or 'intelligent' when applied to technology implies a special property to the technology that clearly does not exist. We have terms such as pervasive or ubiquitous computing that refer to systems of multi-sensorial inputs that are configured in such a way as to provide meaningful output, which are being used in telehealth systems as well as experimentally in University laboratories. These types of system are clearly differentiated from the 'smart' home and in many ways make it look considerably unintelligent.

One of the clear features that has come from the DIRC project into Dependability and Ubiquitous Computing in the Home (http://www.dirc.org.uk) is that 'less is more' or to steel from E.F.Scumacher (1973), 'small is beautiful'. When designing a system for older or disabled people, the simpler the system, the more effective it is likely to be. Moreover, the more chance there is that they will use the system as intended and benefit from it. These systems that are designed are not 'smart', not 'intelligent', they do not think at all. They might process vast quantities of data in a very short space of time, but they can only do what they are programmed to do.

We live in an extraordinary era of change, where it is impossible to keep up to date will all the new inventions and functions of technology that emanate from all around the world. But a worrying aspect of technology intervention is that some people are attempting to make technology be predictive of human behaviours. Systems are being designed with elaborate algorithms that predict actions and activities and then the system decides what course of action to take. This worries me, as this harkens back to the first and second types of people who use 'smart' home technology. The idea is let us see how far we can go with technology so that we can build a truly 'smart' house. The predictive house; the house where you need never worry, or for that matter never need to do anything. The notion of the fully predictive house is like 1984 all over again. It also harkens back to Star Trek! (I know which I prefer). Moreover, can we ensure that a predictive algorithm is dependable and robust?

The answer to both I would suggest is NO! We already are suffering from the 'couch potato' society, in which people do not need to move in order to obtain gratification. The remote on most people's television is a wonderful invention, but it does stop you getting out of your seat and actually engaging with the television set in order to change channels. We can also microwave our food instead of making it from scratch using raw ingredients (which as a result have shot up in price making them too expensive for the people without microwaves to use them).

The second question concerning the dependability of the system is more complex. We all have activity patterns, we all go to bed and all get up afterwards. We all eat and drink. We all get rid of waste products. Most of us attempt to keep ourselves clean and tidy. Most of the adult population have sex (some more than others), and most of this is done at night or in the morning. If you have kids then your routine will be based around their needs and the needs of the educational establishment that they attend as well as the extra - curricular activities they are involved in. Hence it might be easy to build an algorithm that takes most of these behaviours and activity patterns. Simple situations where exceptions to the norm could be built into this algorithm, such that if you had not arrived home by a set time, then the system could assume you are late because you are held up in traffic. Or if you arrive home early one day the system can be programmed to take a number of steps to heat the house up and turn certain lights on and make the food earlier etc. All these are foreseeable eventualities. They are the sort of events that befall anyone in the normal cause of events.

This is fine for the first, second and third group who are not as dependent on the technology to support their life, just their life choices. For the final category these forms of algorithm could be potentially fatal. Older and disabled people's activity patterns are markedly different. There might be considerable rigidity to the format of their weeks, with activities accounted for daily. Having said this, our work in DIRC has also found that when the routine breaks down it breaks down completely. The predictability of the person's routine is dependent on externalities such as friends popping in to see them, visits to the doctor, hospital, shops, garage, friends, relatives etc. Inactivity in the house for a constant period, for example, could mean three things; the person has gone out, fallen asleep or is in serious trouble and needs assistance. So when does the alert get sounded? Should it happen each time there is inactivity, in which case the system would be switched off after the first time the person goes on holiday and never switched on again. Should the system consider a number of factors, such as the front door being opened, activity in the house prior to the time of decision etc. However you consider the facts, what we are doing is trying to contain and constrain the activity patterns into nice pieces that can be easily coded into an algorithm.

But life is not like that.

We are all aware of the unreliability of technology. Although modern technology is more reliable than it was twenty years ago, things still can go wrong. When this happens, it tends to be on a greater scale and costs more to put right. An example of this happened to me recently when the automatic locking system on one of the car doors failed making all the car doors unable to be locked. There was no warning that the car locking system had failed and as a consequence for a number of weeks the car had been left effectively unlocked as each time to lock was applied it would automatically unlock itself, it was pure accident that the fault was traced. When the garage looked at the car the cost of replacing the door lock was far higher tan if it were a manual locking car.

Similarly, we live in a world of built in obsolescence, such that things are designed to fail after a certain period of time or number of uses. Technology in the home is required to be different. Failing technology will annoy the 'Gadgetheads', the 'Glamourheads' and the 'Ecophiles', but it is highly unlikely that a failure will be more troublesome than the cost of getting the devices repaired or replaced. In the home of the older and disabled person, the potential result of a device or system failure is major with psychological as well as potential physical consequences. The highest of these being the resident might die, for example, if the door is unable to be opened.

It is therefore the contention of this paper to address the renaming of 'smart' home or 'intelligent' home technology. So what can or should replace it?

One thing to consider is that we already have a number of words that can replace it. Ubiquitous or pervasive computing sums up exactly the use of technology fading into the fabric of the house and the use of advanced monitoring, but this in itself is a statement concerning the technology and its role. In some ways you can argue that there is an inference towards the person within the name, but it is hard to sustain this argument.

Although, we have suggested four main role for technology the 'Gadgetheads', the 'Glamourheads' , the 'Ecophiles' and the 'Supportive', the first three can be grouped together under the heading of 'technology independent' where as the final category must be the converse, 'technology dependent'. This again is referential to the state of the technology first and person second, hence it might be better to reconsider readopting the already existent assistive technology and non-assistive technology to characterise the two states. One could even go further and add the word electronic to precede both conditions in order to add accuracy to the names whilst distinguishing them from other forms of supporting/non-supporting devices such as wheelchairs and tables.

This paper is not meant to be earth shattering. It is not even meant to be particularly original, but it is meant to try to put to bed the notion of a 'smart' home. As the title suggests they have failed the intelligence test, the terms are obsolete like much of the technology. In order to dynamically extend our thinking and understanding concerning technology, the home and people, we need to ensure that we are all speaking about the same thing, using the same language. Hopefully this might paper might begin to open this issue up for debate.

©Guy Dewsbury 2003 smartthinking.ukideas.com