|
The Challenge Network www.chforum.org |
|
|
This paper describes an application which has potentially vast consumer appeal, and which draws on the unique features of third generation (3G) mobile telephones. The examples are primarily British, for which we apologise to international readers.
This paper is posted by Datafreeze® Limited as an aid to debate. Concepts in this text are subject to ongoing patent applications and Datafreeze® Limited asserts intellectual rights over the entirety of the ideas here presented.
A virtual network is a layer of linked telephones, collectively representing a cadre of colleagues, friends or those with a set of special interests. Any one telephone can support several such networks as independent layers, all of which are permanently active. This creates an infrastructure on which users can build, so as to represent themselves and their interests - including corporate or business presence - through such networks. Please see the concluding two paragraphs in this section for the commercial significance of this.
Why people would want such a facility is, perhaps, too novel to easily summarise meaningfully. However, an example of a session is attached, overleaf. The first section of the text - "the concept" - explores both the facility and its many users and uses.
The technology to realise this concept demands the high bandwidth "always on" characteristic of 3G mobile telephones, plus a server infrastructure which only 3G can easily support. However, it is not otherwise particularly difficult to realise. There are three key issues that will need resolution: the interface(s) and thus the physical form of the mobile telephone; the security of the networks and the requirement for geographical information about the location of the handset. The last two of these are solved by the technology being developed by Datafreeze® Limited. Datafreeze® Limited holds patents pertinent to this focused issue as well as to the generality of this concept.
The new 3G mobile telephones are in need of distinctive product offerings, which draw on their unique capabilities. Fast-developing applications in communications, such as conventional mail, the Internet or even groupware used on intranets, have almost always allowed their users to reconfigure and otherwise build upon the basic infrastructure. Offers which cannot be configured by users - for example, channelled television - do not generate such opportunities for independent, bottom-up growth. Such products survive or fail on the virtue of their content; and revenues derive largely to the vendors of content, and not to the infrastructure providers.
The openness of virtual networks is a crucial feature for their acceptance as a pervasive application. In commerce, for example, they enable knowledge management, and revolutionise project and risk control. They create a new way to be social - to be a family or a group of friends, for example - and they provide a platform on which all manner of added value activity can be based.
John glances at his telephone as he takes his seat on the subway. There is a great deal going on, and his interface is alive with activity. It is chiefly driven by his membership of the True Blue Beer Club. He runs his finger over the interface, and one of the prime drivers of this excitement, his friend Bob, pops up. Bob's picture is surrounded by a cluster of icons showing that he is in a pub, that he is with friends, that he is enjoying himself and that he is communicating madly with everybody can think of. The pub attempts to chip in with directions and an invitation tailored to John's public persona. John is tired and he does not feel like socializing. He slides his finger across the interface to bring up his family network.
All three children are at home, but his wife is out shopping at the local supermarket. Tenderly, he strokes her icon on the interface. Detail blossoms around it. He glances nervously at his neighbour in the underground, because this is private matter. He taps the image representing his wife. A familiar message appears: "John, I'm busy. I'll call you later". John grunts to himself. The supermarket bids for his attention, as does her car, which needs petrol, a service and a new license. He brushes both of them aside.
He moves his finger across the interface, resting it on each of the icons representing his children. Two of them are in their bedroom, presumably watching television, although they are also interacting through their mobile telephones with their friends. John smiles to himself.
Anna, the third daughter, has entirely disabled her self-reporting to the world at large, and has cut off as much as possible from her parents. This has not stopped her running up a phone bill as large as the rest of the family combined. She is locked in a one-to-one with her boyfriend of the moment, and has been for the past three hours. Her credit card nags at John - it is due for payment, and Anna has put a permanent message against it asking for an increased allowance. John pays the debt, cancels the standing message and asks her to speak with him when he gets home. She responds with a beep of acknowledgement: not even a written message. John fumes.
Worry about Anna is brought to a halt by a message from the office system. It asks whether he wants to disclose a set of documents to a colleague. He agrees. A second message invites a quotation for five minutes of consultation, to begin immediately. He enters a high bid, which is immediately accepted, somewhat to his annoyance. He fits his earpiece just as the team materializes on his interface: same old faces, but with a dense cluster of in-house live links running out from the icon representing the technical department. He notes the presence of an outsider. A tap shows that this is a senior figure, representing a major customer. John sits up straight and changes his contingencies to exclude distractions, such as Mary's expected call-back. He also checks that the presence of an outsider has put his system into the proper disclosure mode. The meeting begins as the train thunders through the New York suburbs.
Virtual networks attempt to shadow our normal way of living, but to do so whilst eliminating the restrictions of geography. This is a new concept, so we will approach it in stages.
The individual members of the various networks to which we belong come into focus for us as circumstances demand. Such circumstances may be a one-on-one interaction, or may be group-based. How we present ourselves to any one group may be different from how we behave elsewhere. At times we wish to be at the heart of a given group, and at other times we wish to exclude it. The status - activity, business - of a given group may attract or repel. That is, the group dynamics, or events built into how we interact with it, may affect our communication with it. Excited play may draw us in, or make us cast a parent's eye over the proceedings.
How might these features of everyday life be embedded in a 3G telephone systems? Plainly, through a structure which mirrors these social processes.
In telecommunications terms - but in not way different from daily life - the individual has to be tied into a set of networks to which they supply information about themselves. They can tailor the appearance of this information, and they can set it to be as revealing or as secretive as they wish. Two key components of current, as opposed to general personal information are, however, where they are, and in what form of activity they are engaging. How this information is collected, presented, managed and secured is, of course, crucial, and this is discussed later.
|
Let us begin by confining this discussion to just one network. The network is populated by Millie, Billy and Willy. Exactly what Billy and Willey might see when they glance at their interface is discussed in the next section. |
This said, let us assume for the moment that they would see an icon which they knew to represent Millie, with a "house" symbol near it and a low activity level. This might, perhaps, mean that Millie is asleep, at home. (Whether Millie would want to declare this, and what she might do about it is, of course, discussed later.)
People would, however, customise their symbols (busy, at work, shopping...) to meet their needs. High resolution positioning systems would allow the automatic choice of icons to represent specific locations - my office, this or that shop - and commercial premises could be organised to interact with telephones to deliver this level of precision. Icons could, therefore, become extremely elaborate, given high bandwidth and a display with good resolution.
Each individual could choose to present themselves as a cloud of pictograms of their choosing, expressing as much or as little as they chose about themselves. Third party icons (company "business cards", affiliations "Arsenal!") and, perhaps, tools such as credit cards could also be displayed. Some of these icons could be active, in the sense of hyper-linking, or perhaps opening to reveal yet more information. Commercial applications of this, including advertising, are considerable. The business aspects of this are self-evident, impinging on knowledge management, project management, safety and security. Public sector services could generate extremely important applications for their employees.
The presentation of this - the ability to zoom up to the networks level and down through several layers to individual details - will be a necessary feature of the interface, probably best implemented through a high resolution, touch-sensitive screen.
Readers may object that nobody would want to declare such things to strangers. That is precisely what secure virtual networks are designed to avoid, however, and users have absolute control over who they include in their network(s) and what they show them. Indeed, they may want to show one thing at one time of the day or situation, and quite another when their status alters.
It is clear from very many surveys of social attitudes that many people long for community. Indeed, this longing is the basis for the SMS takeoff and for the closeness with which come people guard their mobile phone. The key issue is who is included in the network, and what you wish to reveal to this particular group. Millie may be close to Willie and Billy, but may describe herself quite differently to the network for whom she is primarily the district midwife.
Equally, how the group is presented - what it declares and to whom it declares it - is also open to design. A network of childhood friends might be envisaged as a ring of icons, with lines connecting them to show who is conference-calling whom, who is calling outside the ring and, of course, who is inactive or otherwise engaged. The location and activity status ("hanging out", "shopping") of each would be accessible. As we have already suggested, locations such as cinemas would be matched to locations of telephones, and so not only would the group know that one or more of their friends were at the movies, but they could know what they were seeing. They might link up to share the experience. The cinema might ask the child to review the movie for their circle and give them free tickets to do so; and so forth.
This pattern would not, of course, do at all for a complex organisation. It would need more subtle layers of access. Further, it would make heavy use of 'contingencies', which are discussed below, under implementation. In brief, however, this means that users can easily set their telephone in pre-established states of declaration, from the opaque, through various forms of minimalism to full linkage. They can do this at will, during particular time of day or when they are in specific locations.
Now let us introduce the possibility of overlapping networks. The network established by Millie for her children would be completely different from that which contains Billie and Willy, or her work-related network.
An individual sets up - or is accepted into - several networks. The information which is presented to each of these, and the nature of the presentation, is potentially quite different. A company network might have a standard appearance, and might contain the equivalent of a business card that could be included by the individual in more general contacts. It might offer a diary, access to a corporate LAN and the like.
The user - Millie, in this case - can pop in and out of these layers through her mobile interface, which is also her telephone, her PDA and the nexus of all manner of other functionality.
She has complete control over this, subject to the limitations imposed by the networks to which she belongs. However, the networks themselves - or particular individuals, processes and way stations - can also press themselves forward attracting her attention to them.
This would happen in any number of ways, something that network programmers (once again, see below) could establish for themselves.
Instances might include:
The uses are limited only by user ingenuity. It is essential, therefore, that the user's interaction with the overall system is as flexible and as open as possible. This flexibility extends to how the individual interface is presented, and how a given network is to appear and operate. It covers individual membership of networks. It covers how individual mobiles are to interact with such membership: with what we have called 'contingencies'.
These issues of interface are discussed in more detail in the section which follows. The key point to hold in mind is, however, that it is the users (individuals, companies, bureaux, third party software vendors) that undertake this development. Virtual networks must not be stereotyped, or straightened by the service providers, or they will never grow to meet their potential. The analogy with the Internet is self-evident. If all web pages had been written by PTTs, and if users had no influence on what went on in the Internet, then it would have been still-born.
However, as the next section shows, it is a mistake to see the mobile telephone as the equivalent of the Internet terminal, with its embedded browsers and other Internet-handling software. One of the advantages of 3G is that the processing can be done remotely, and the display of the results of this alone handled on the handset.
This and related topics is, however, discussed in the next section. The crucial point to bear in mind is that managing the structure - network appearance, network membership, permissions, contingencies and so forth - are not handled by the handset, but through secure internet connections to the virtual network server.
Technically-adept users might want to set up their telephone through its handset, much as they do today. The next section explores the reasons why it is not a good idea to do in this much more complex arena. It seems more appropriate to do this through a separate mechanism, based on the personal computer and operating through the Internet. This is set in its context in the next section.
In general terms, however, the Internet-based configuration management tool will allow the validated user to operate securely at three levels.
First, they can apply to enter existing networks, or set up their own network. Where they are the network's web master, they can define its generic look and feel: layout, fonts, colours, sounds, animation and other web-like paraphernalia. Given the now huge skill base in HTML, it would be wise to consider whether it, or some subset of it, could not be used to at least designing this interface. This issue is discussed in depth somewhat later in this section.
Second, users can define their own presence on each network. This occurs in two blocks. One of these defines the extent of the information that they are prepared to declare, and 'contingencies' around this. ('Contingencies' are discussed in a moment.) The second block of declarations is concerned with how they want to symbolise this information. They may want a picture of themselves to stand as their icon. They will need to define where (geographically) locations like "home" and "office" - very much etcetera - are to be found. They will define some alias icons - "gone fishing", "over the moon" - for when they wish to be private. Assorted contingent icons - "I am lonely", "Do not disturb" - can be set up. One could envisage a large industry supplying these static or animated, personalised, customised tags.
Third, the user can set up "contingencies". Some of these have already been met: the user is at work, or at home, requiring privacy or seeking fun. Each of these modes places limits on the way that the user is represented to the various networks in which they participate. Each can be evoked by turning them on and off directly, by having default times when certain things apply - "sleep" from 23.00 to 06.30, "work" Monday to Friday, 07.30 to 18.00 - or location based switched, when presence in the office triggers "work". All manner of fuzzy logic combinations can be envisaged, as can third party bolt-ons.
It is important to realise that the mobile telephone is a part of a system by which the user interacts with their virtual networks, and not the only system. The telephone is not a limited, shrunken personal computer. Much that it displays is computer by a server and shipped to it, and the information on which this computation is based is taken from many distributed systems, including all of the telephones with which it is linked, with business intranets, with contingencies which the user has set and with applications that third parties - shops, news casts, commercial offerings and so forth - are also trying to press on it.
This sounds extremely complex, yet - excepting the handset interface - it draws on existing or planned infrastructure. The basic framework looks as follows:
There are two novel "big system" entities. On the left, we find the virtual network server. This holds the user's details and the network details. What the user sees on their virtual network interface on the handset is, therefore, computed in this system and served to the handset through the PTT network. This system takes an overview of the status of all of the networks that it serves, and updates the relevant handsets with the information that this entails, as well as responding to demands from the individual user. Although not shown in the diagram, it also interacts with proprietary virtual network servers - that is, those of companies, semi-commercial affiliations ("the RSPB wildfowl network") and other organisations. It passes these proprietary servers information in respect of the relevant handsets, and it receives instructions from them as to communications which should be sent to these.
Users can make adjustments to the way in which they are represented in the various networks in which they participate. "Webmasters" can alter the look and feel of a network, its permissions, 'contingencies' and so forth. The proprietary systems - such as corporate intranets - will allow individuals limited discretion, much as companies allow individuals to tailor their screen interface and set up personal web pages. However, the corporate "webmaster will create a standard design for the network, and set permissions and other features of it. There will be a considerable cottage design industry, much as web sites are designed by third parties today.
On the right of the diagram, we find the second novel element, which is the Datafreeze® system. This handles encryption and location-based information. A summary of the Datafreeze® project is given in the box, below.
Datafreeze®Both EU and US authorities have mandated that mobile telephones will be equipped with GPS systems that can report where they are being used in the event of emergency (911, 112) calls being placed on them. This is a pure regulatory cost to handset manufacturers, who are resisting the legislation. However, Datafreeze® offers a means to turn this into an advantage. The technology uses GPS data as the basis for on-demand, real-time and low cost secure encryption between two such equipped telephones. (Static telephones can be equipped with a device with plugs between them and the wall socket.) Encrypted calls are recorded securely in Datafreeze® installations, where authorized individuals (the telephone owner, be it an individual or a company) can delete, archive or download a record of that call; or simply allow the encrypted record to expire and be erased. All calls are evidentially watermarked with time and location. More can be found here. The Datafreeze® project is a relatively inexpensive one to bring to fruition. The combination of location-based identification and secure communications is, however, essential to the establishment of 3G virtual networks, where security and assured information about the corresponding handset and its location are central. |
There are two design aspects to the interface which merit general discussion. One is the underlying software capabilities which it offers, and the other is its physical design. We are not experts in the later, but we offer some views in order to point up the need go beyond traditional approaches to handset design.
There are two fundamental interfaces with the virtual network. The first is interface is the PC-to-Internet system that allows the user to configure their system, create a network and so forth. This has already been discussed in outline. The second is, of course, the mobile telephone itself, which we will discuss in a moment.
Both systems could, technically, be collapsed into one, but at the cost of greatly complicating the handset. Centralised systems management means that change management does not have to worry (so much) about embedded hardware or legacy software. As already noted, much of the processing that appears to be handset based will, in fact, be handled in the network server. Refined calculations would be needed to decide which aspects are best served by local processing and which by uploading.
Let us begin with the supporting PC-to-Internet system.
In principle, this could be handled through the telephone handset. However, it is notorious that users do not read manuals, and that they are often confounded by the complex interfaces that are bundled into a few keys and small-screen menus. (Over three quarters of 'veteran' VCR users had not learned how to program their machines, according to a survey made in 1995.)
It is, however, essential - crucial - that users set up their information declaration and networks (or have them set up for them by others) without making revealing or costly errors. This means that it is essential that the tools are displayed in the easiest and most open manner. A PC-based interface is far easier to address - indeed, for anyone over 50, to easier see - and so to understand than is a complex handset and small-screen interface. The software needed to create wizards and video training are easily available. The Internet lies within easy reach for downloads, advice, third party help.
Virtual networks will generate many business opportunities. In this area, bureaus and other small business services can work with the user to deliver what they want, perhaps as Internet businesses, perhaps as physical presences. The draft can be traded back and forth on the Internet, and then instantiated on the server when the user is happy. However, the interpretation of this choice will almost certainly be server-based. If, for example, a user joins a network which has an established design, then their file on the server will simply access the relevant information. Data computed for export to their handset would then use this as the equivalent of a style sheet; or perhaps HTML template.
Setting up one's identity on the net needs to be intuitive and simple. The issue is not only one of confidentiality, but also of appropriateness. That a users supports Arsenal is not usually confidential, but it is always inappropriate as a part of their presented persona in some area. By contrast, that they are called Fluffy by their wife is a matter that is highly restricted, as is - in a different domain - the fact that they are a detective chief inspector.
This suggests a matrix. Along the top, we have items that are a part of "who they are". They can add a large number of items to this list, but most will choose at most a dozen. A few of these - their telephone number, their name - are basic. Some items come from a corporate base (that they really a member of the Metropolitan police) and others are discretionary ("Fluffy").
Down the side of the matrix we have the networks to which they belong. There is one default setting: public, which is innately minimal in what it allows to be said. Then there might be more: the job - with assorted subsets of it created by the job's netmaster, then domestic-home (wife) and domestic-home (children), Real Ale Group and Tesco shopping. And so forth.
Note that "Arsenal" is both a self-declaration and also a separate network. The credit card is broadly declared, and validated by the issuing company, but is details are accessed only on user say-so. RSPB alert! is an entirely passive "pager" application.
The intersections of the columns ('what I want to say about me') and the rows ('the networks in which I say it') start out empty. By clicking the cell, the user can enable it. Nickname 2 - "Fluffy" - is ticked in "my lovely wife" only, and will only appear there. The Job-related cells are probably ticked by the netmaster - and much more complex than shown here - but some of them are set on user discretion. However, the most public layer - the business card, the warrant card - would be accessible to all other telephone users, whilst the "Senior staff at Sunhill police station" net would be available only to immediate peer colleagues.
We have already discussed "contingencies". Any one row could be set to default up a layer, or to a nominated layer, when a contingency is evoked. That is, if the user evokes a contingency - presses "Cloak me" - then all calls are pushed up a layer, or to the Public setting, as they choose. Their presence on the networks is similarly modified.
The issue of contingencies of time and place offers powerful opportunities, but also need careful interface design. That is, the non-work aspects of the user's networks are strongly de-emphasised during working hours, or perhaps when the user is in certain locations. (Recall that how a user will be presented depends on which network is looking, and that it is not necessary to call or be called in order to be a presence on a network.) In project control applications, absolute strictures on location and communication are imposed. Equally, off-duty, "fun" aspects are emphasised, and potential entertainments are served up.
It may be something for professionals to set up, at least until and expert cadre emerge. Plainly, there is much to be worked through in this area, striking a fine balance between simple default modes and access to the "Meccano" of the system that will allow new industries and ways of operating to arise.
The mobile interface is, of course, where the value of this system is delivered. The nature of this value places certain physical demands on the interface, and requires a standard set of software capabilities. We do not intend to major on the physical nature of the interface, but a few words may be appropriate.
The physical interface. The user interface can, of course, be created in many ways. Mobile telephone manufacturers prototypes for 3G handsets still tend to maintain the traditional "mobile" look, whilst slightly increasing screen size. This may be a mistake, but it is clear that the successful design will emerge from the competitive struggle. Our own assessment is that the mobile will be physically split into an extremely light speak-and-hear component, and a separate "see" element. (It is hard to look at something into which one is speaking in a public place.) Connections between the two would be wireless, subject to battery life in the headset. The two would plug together when not in use, equalising battery charge.
The "see" element does, however, absolutely need a relatively large area of high resolution screen, operating to at least the standard of a computer monitor. This is required in order to support the information density which the virtual network would be capable of delivering. How the user would interact with this is open to debate. Options range from touch-and-voice feedback, where the screen can be configured to any command interface, to more conventional WAP-like keypads, using the 'folding baguette' design. New flexible organic LED screens offer the option of scrolling or otherwise extending quite large screens from pocket-sized devices.
This said, it seems likely that a small number of keys - chiefly numeric, but including unconventional ones such as "encrypt this conversation", or "change contingent mode" - will be coupled to a touch sensitive screen. Keypads and other less frequent applications -games, for example - could be displayed under this membrane on demand.
The major feature of the virtual network - the ability to oversee the net, and to zoom into elements of it - is ideally suited to a touch screen. A simple and intuitive manner of selecting amongst one's contacts would be arise if the area under a fingertip simply expanded, as under a magnifying glass, and then subsided when the touch was removed. However, as we have already indicated, time will tell and several variations will almost certainly coexist for different market segments.
The software capability. We have suggested that the bulk of the processing should be done centrally, and controlled both by events and by style sheet and HTML-like instructions as to the appearance and behaviour of the interface. The outcome of this are transformed into graphics primitive instructions, much as they are in a standard computer, but mapped onto the specific handset which the user holds rather than the PC graphics card.
The outcome would constitute the signal stream that actually goes to the handset. Bundling a browser-equivalent capabilities onto a handset would be costly and daunting for battery life, and add no obvious functionality, or at least not where bandwidth is plentiful. Screen refreshes would not need to be particularly swift - certainly not at PC rates! - and particularly not if animation and other repetitive tasks could be handled in the handset itself. This would also apply to simple graphics roll-overs and reveals.
If the major part of the processing overhead is to be placed in the network server, then the handset needs only to have a limited repertoire of graphics primitive handling and touch pad management. The benefits are longer battery life and lower cost. The downside is increased bandwidth requirement, and a stupendous central - or more probably distributed - computing effort. This is a topic that demands in depth investigation.
This said, the graphics interface software needs to be able to handle:
Multi-font and format text
Static vector graphics: polygons, floods, gradients
Dynamic vector graphics: scale, rotate, translate
Layer clipping
User-generated screen interrupts
HTML form equivalent input.
Perhaps excepting the last, these are standard library components. Some manufacturers may wish to use voice-operated software, roller balls, extended conventional keypads and the like, so the information fed to and derived from a given handset will depend on its make. This is an identical problem to that of computer peripherals, solved by having assorted drivers. What is needed is a basic standard to which all must adhere in building machines and applications.
It is clear that virtual networks are a natural application for 3G mobile telephones, drawing on its unique functionality. The basic architecture needs only one novel element, the Datafreeze® location and encryption-providing system.
There is, of course, considerable work to do in respect of systems design and integration, but the component parts are available off the shelf and the concepts are familiar. Much thought needs to be given as to the correct user interface. In parallel, the issue of where computation is to be done also needs definition.
The possibility of linking virtual networks with the Internet in some way is extremely attractive. This would not, however, be to display web pages badly. Naturally, the devices will handle e-mail as a matter of course. Rather, there needs to be a mechanism whereby existing database structure can be accessed transparently (to them and to the originator) through distinct interfaces. That is, the way that RailStuff.com would be presented on the mobile handset would be tailored to that application, but the underlying database would still be addressable by it. Aspects of XML are well-suited for this purpose, of course, and the technology was developed for the WAP program. However, sites that are not so equipped should be invisible to handsets.
The technology does not, therefore, present insurmountable problems. The crucial issue is consumer acceptance of what is, ultimately, a socially- and not a technology-oriented service.
Much effort will need to be put into the generation of user understanding and enthusiasm for a system which, potentially, has a steep learning curve. This is particularly true of self-declaration, and the design of new networks.
It is likely that commercial users will lead in the complex area, and that school children will be the early adopters in simple, pre-designed networks. The communication of the concept to the public needs to be managed in a less chaotic way than characterised public understanding of the Internet: that is, by dread followed by an irrational gold rush.
| To top |
Links to Interactive Trek Guides sites for Peru and Nepal.
The trek Peru web site. The trek Nepal web site.