The Digital Home - The Grid Approach

For a few years now one of my research domains has concerned the Future of Home Networking. My interest in this area of technology was I guess, initially sparked as a geeky 10 year old hacking away at my ZX81 with an imagination fuelled by Lucas, Roddenbury et al.
So when I was asked to deliver a keynote at last week’s inaugural Wireless Grids Research Consortium Meeting on the topic of visions for the use of Grids in the home environment, starting with the mental exercise of deconstructing the Star Trek ship's computer seemed only too natural.

From HAL to the USSS Enterprise and beyond, the intelligent Star Ship Computer has become a cliché staple of Science Fiction writing and film. Be the question ‘Computer, what deck is Mr Spok currently on?’ or ‘How long before that planet explodes?’ the Star Ship computer has the answer.
However the computer doesn’t ‘know’ all the answers, it simply ‘understands’ the question, ‘knows’ where to find the information required to present a logic answer and 'understanrds' how to manipulate the physical devices which can supply information about the environment.
This glib (and over long) sentence is of course ascribing simplicity of concept to something complex in practice, but for the sake of illustration consider the oft utilized plot device of a stricken, powerless Star Ship in decaying orbit around a planet. The question ‘Computer, how long before the ship burns up in the atmosphere?’ would more than likely result in a prompt answer of the form ’37.2 minutes!’
If we examine this we can see that in order to answer this question the computer must:

1. Understand the question and also:
1. 1 Recognise the questioner
1. 2 Recognise their context
2. Locate information about current speed from ships sensors
3. Locate information about planet mass from external sensors
4. Calculate the mass of the ship from ship record and cargo log
5. Find appropriate equations for planetary physics from some knowledge bank
6. Apply all the information to the equation in order to calculate the answer.

We can also see here that even the method for calculating the answer, given the other data sources, can be discovered through an information source. The intelligence therefore lies in being able to aggregate the correct information and compile it accordingly.

Now, this may seem ridiculously geeky but it allowed me to make an important point. Home networking technologists are often so concerned with the networking aspect that they neglect to investigate all but the most trite and obvious use cases, such as moving files from one place to another or sharing an internet connection that the notion of what this new 'network/device gestalt' is capable of is often overlooked.
I've lost count of the amount of times some techno savvy type has piped up with an assertion that building a home network is easy and how simple the Windows Network Setup Wizards are etc. Indeed not being alien to technology myself, my own home has boasted some form of computer network for at least 8 years. However, these assertions always fall into the same traps of :

1. Assuming that the average home user has any interest in building (and more importantly maintaining) a network
   
2. Assuming that computers are the only thing that needs networking.
3. Assuming that is other devices are to be networked it is so they can be exploited by a computer in some scriptable liner process
   

Beyond the expensive home automation and entertainment integration systems available at the high end market (which are very cool) I would go so far as to say that there is no such thing as a home network at this point in time. Sure, we can scale down an corporate LAN but to my mind any system that requires someone in the home to take on the role of a network administrator can not really be considered a 'home' network.
There is the argument that this could be set up by a third party expert much the same as the electrical or plumbing networks and this would be fine if not for the small matter or reliability. I have not had to touch the core electricity or plumbing networks in my house since installation, yet my computer network needs weekly attention.
I long since coined the term 'the family CTO' to label that person existent in many extended families to which all others turn when computing problems arise, but it is not really viable for network and technology suppliers to rely on their possible existence or assume that there is any interest in domestic systems integration. It is not a new idea that things sell better when they are easy to set up, use and above all understand.

So, getting back to the Star Trek Ship’s Computer, to me the point of home networking goes beyond the relatively simple aspect of connecting devices together- rather the point is to enable the end user to do something previously impossible but to allow them to do this in a simple way. By connecting the computers to other devices with (or potentially with) communications capabilities - games consoles, cell phones, stereos, speakers, TVs, white goods and even the family car - we can start to construct a network of capabilities (resources, in the Grid computing parlance) and think about how abstracting device capabilities from their physical boxes by way of the network generates new possibilities.

If we can some how place the intelligence for connectivity, identity, security, information retrieval and user orchestration into the network we can begin to extract values from the devices found in a home on a level greater than the sum of their parts. This may need some kind of hub, though the notion of making the network itself the ‘hub’ is personally more intriguing.

Maybe this is all science-fiction, I admit at least that this is a longer term vision, but through some of my work with the Wireless Grids community I am starting to see some real possibilities.