Position Paper for the WWW9 Workshop:
The Web and Mobility - Information Design for Mobile Web Services
IS Technology Lab, mobilkom austria AG
Obere Donaustr. 29, A-1020 Vienna Austria
A look at current buzzwords in the web and mobile markets should be enough to set off the warning bells in the heads of those who really understand the technical implications. Everyones talking about anything, anywhere, on any device. Services arent just services theyre optimized, customized, contextualized, personalized and localized.
And what about portals ? Nearly everybody has a portal. Anybody whos anybody probably has more than one web portals, wap portals, voice portals, portable portals. If the trend continues, what everybody will really need is a portal for their portals.
If all that sounds provocative, thats fine, because it was meant to be. Provocation in small doses is a good catalyst for thought. The things that should be thought about are not only deployment of innovative technologies, but more important, how they can be unified and optimally exploited for the benefit of all concerned for manufacturers, network operators, Internet service providers, software developers and consumers.
mobilkom austria is the leading GSM provider in Austria, with a penetration rate of 54,6% and 52,2% market share. In keeping with its goal of maintaining technological market leadership, mobilkom austria is committed to the continuous development and implementation of innovative ideas and technologies. One good example is mobilkom austrias A1 PocketNet, launched in 1997, which offers subscribers mobile and fixed Internet access in Austria, as well as mobile access from most European countries and many other countries around the world. A1 PocketNet also includes innovative web-based services, such as ÖBB Ticketing, a service for railway tickets which can be ordered per WWW, WAP or SMS and are issued as short messages and billed to the customers mobile phone.
IS Technology Lab is constantly on the lookout for innovative technologies, focusing our efforts on finding those which are the most interesting for mobilkom austria, conducting lab trials, and, if proven suitable, recommending deployment of those technologies. Were also working towards cultivating relationships to manufacturers, as well as to other GSM operators and technology labs.
IS Technology Labs WAP lab trial began in September 1999. The experience gained from the the lab trial proved to be invaluable for the deployment of WAP. Activities included working in the WAP project team and also communicating the knowledge and experience gained during the lab trial to other departments, as well as to business partners. WAP was launched in December 1999, making mobilkom austria the first to bring WAP on the Austrian market.
During the WAP lab trial, we also began investigating and testing solutions implementing XML and database technology for personalized portals and cross-browser, device-independent access. We are now in the final phase of our WAP lab trial, which consists of developing a concept for WAP access to corporate Intranets.
Corporate GPRS is presently being tested in our lab as well, and solutions for corporate access in general is an important topic for us. Were also currently conducting lab trials of voice over IP, and a speech recognition trial is in the installation phase, to be followed later by text-to-speech. Lab trials of position-related systems are scheduled to begin soon.
The Synthesis of Technologies for the Optimal User Experience
Some might find it strange that a person coming from the technical side of the fence such as myself would voluntarily propose discussing the issues from the user perspective. However, learning how to deploy the single technologies themselves is not the real issue weve all done that. There are RFCs, protocols, specifications and white papers which spell things out. Everybody abides by the rules and everythings more or less fine.
Providing mobile access alone isnt enough though. The convergence of the traditional web and mobile, with a pinch of personalization, localization and customization thrown in for good measure, requires a new way of thinking. Its not only about technology anymore, its also about philosophy.
Due to the complexity of the various technologies that will most likely, and ideally be implemented parallel to one another, it seems that a good way to get the right perspective - the big picture - is to look at things from the end-userss point of view.
The average end-user isnt interested in the fact that his favorite service may have been implemented using personalized data, customized with information from a new database-profiling system, and that the localization system is providing his current whereabouts. He doesnt really care if its push or pull either he just knows that he gets the information he needs or wants, when he wants it, where he wants it and how he wants it.
From that perspective, we at least can see where we should be going, although it may still be unclear exactly how we should get there.
End-Devices Then and When
Perhaps taking a look at where we came from will help us to find out how to get where were going. Dividing the end-devices of the converging worlds into two major categories of aural devices and visual devices provides another perspective.
In the beginning, there were aural devices, such as phones first fixed line phones, later radio phones and mobile phones - aural devices without any visual capabilities. Later, technologies such as SMS and WAP made it possible to add visual properties to these devices. Someday soon phones will have visual capabilities similar to those of PCs.
There were also visual devices, such as PCs. In the beginning, there were desktop PCs. Aural properties were then added to PCs by the use of soundcards, speakers and microphones. Later there were portable PCs, laptops, notebooks, handheld PCs and personal digital assistants. Now even wearable PCs are evolving. Most of these devices now have aural capabilities as well as visual.
However, even with the converging capabilites of end-devices, there are still drawbacks in both worlds. Mobile devices still have limited displays and memory, as well a limitations for user data entry. PCs and handhelds have good displays, sufficient memory and more practical user-interfaces, but, particularly the smaller, more portable devices may still be weak in regard to the aural capabilities.
The Technical Aspects
Considering only the end-devices is not enough to get a real idea of all the issues that must be dealt with when considering the convergence of the web and mobile worlds. Lets think for a moment about bearer technologies.
From the users point of view, the transport media or bearer is not of the utmost importance, save for the requirement that it be as fast and reliable as possible. Included is a short overview which will hopefully provoke some new thoughts about some of the aspects which should be considered.
TCP/IP is the bearer for the web, in the beginning over fixed lines and/or satellites, now also over optical fibre, copper, electrical lines, or just about anything that one can imagine.
CSD (Circuit Switched Data) has provided the mobile world with mobile access to web-based services, although, before WAP, not with a handset alone, but used in combination with a handheld device. The main drawbacks were, and still are, long call setup times and data transfer rates of only 9600 Baud over GSM.
GPRS (General Packet Radio Service) is also a bearer which will provide mobile access to the web world, with much better possibilities regarding bandwidth and speed than CSD. GPRS can be used as a bearer for WAP- and/or web-based services.
SMS (Short Message Service) is the only one of these bearers that is listed below as a user-interface as well. The main advantage of SMS is that, if not immediately deliverable, it can be stored and delivered later. SMS is widely used a bearer for push services to mobile devices. SMS is the bearer for SAT applications and can also be used as a bearer for WAP.
USSD (Unstructured Supplementary Service Data) uses, like SMS, the SS7 signalling channel. However, since it is session-oriented, there is no channel reservation time per message and there is no "store and forward" delay, resulting in much shorter response times. This makes USSD a good choice for interactive applications.
Now throw all of these bearers in a hat, together with the current market penetration for each, as well as their respective projected market penetrations for the next few years. Is it possible to pick just one or two, and decide to provide services only for those, especially knowing that one user may want to access the same services over different bearers, depending on whether hes at work or at home, or depending on which end-device he happens to be using in the moment ?
An overview of the user interfaces from the convergent worlds may also shed more light on the subject of making decisions about service platforms.
HTML is the language of the web a relatively simple and yet efficient tool for creating usable visual user-interfaces.
SMS took the first steps to bringing text to mobile phones, but being a text-only interface for devices with limited capabilities for data entry, isnt especially user-friendly. In spite of that fact, SMS enjoys great popularity. This is a sign that mobile users want/need visual capabilities.
SAT (Sim Application Toolkit) took things one step further, bringing a more user-friendly interface than SMS, but not all handsets are SAT-enabled. SAT applications are also dependent on the memory capacity of the SIM card. There are also SAT applications which function as WAP browsers on non-WAP-enabled handsets. SAT also includes some functions for interaction with the SIM card on the handset.
TTML (Tagged Text Markup Language) is Nokia's protocol for presenting information accessed from the Internet on GSM phones. The user-interface has a similar look and feel to that of SAT and WAP.
WAP is bringing web-oriented user-interfaces and the web paradigm to mobiles using protocols optimized for small bandwidth, displays and memory. WAP 1.2 will also include functions for interactions with the handset, as well as push functions. WAP seems to come closer to the goal, but the limited display and data entry is still a problem.
Speech recognition and text-to-speech seems to be, especially for phones, a natural user-interface. It could also possibly help to bridge the gap between the two worlds, web and mobile. One main advantage of speech-driven user-interfaces is that they are hands-free and also accessible for the blind and physically handicapped. The quality of the technology used has advanced greatly, and establishing standards for speech-related markup languages will be helpful in making speech-driven interfaces a realistic possibility.
Now that weve discussed transport medien and user-interfaces, lets look at another aspect of the evolving technology. Providing information that is relevant to the user, based on his interests and his current whereabouts will always be advantageous, not only for solving technical problems, such as bandwidth issues. The users time is also valuable and considering the flood of information, the user may easily become accustomed to specialized information, regardless of his whereabouts or which end-device hes using at the moment. There are various possiblities for filtering information for end-users.
Location-specific information and services for mobile access is currently a hot topic. Privacy issues regarding location information have not yet been broadly discussed, but as long as the user is informed that he will be located when using a service and he consents, there can be no disputes. The question is, if its sufficient to be able to provide location-specific services only for mobile access ? Keep in mind that system decisions will be influenced by the answer. Perhaps the end-user will want location-specific information not only if hes accessing a service over his mobile phone, but also if hes accessing it from his car, from home or from the office.
Context-specific information is another way to enhance the user experience. When ordering products or browsing online, most users will appreciate being informed about related products and/or subjects interesting for them on the basis of the current context. Especially in the case of mobile access, where bandwidth and display are still limited, presenting specific information to the end-user is important.
In the GSM world, we have the possibility of identifying the end-user in most cases on the basis of his CLI (Calling Line Identifier). In the web world, its possible to identify the end-user by way of HTTP-cookies or, depending on the degree of security required, with user/password or certificates. Once the end-user has been identified, it is possible to present him with information based on personalized profiles. In cases in which the user connects to a service over more than one source, for example over an access server for WAP, then over a portal, and finally to a service or website, where the profile data should be stored is an open issue. Wherever it is stored, personalized information saves time and bandwidth, and enhances the user experience.
Technical Approaches to Promoting Synthesis
Not all services make sense over all transport medien, but the ones that do should be accessible over multiple bearers, or a choice of bearers.
User-interfaces can benefit from new markup languages, some of which are in the development stages, not only for visual user-interfaces, but for speech-driven user-interfaces as well. It seems that most of the issues regarding user-interfaces can probably be solved using markup languages. Therefore, the global solution for user-interface convergence will probably be based on XML. Important is the use of worldwide industry standards rather than proprietary solutions. Automatic browser and device detection combined with XML can be a powerful tool to take us one step further towards any service on any device.
Regarding the platforms and services in general, important is the use of open systems which can be integrated with other technologies. Integrated service platforms which use the same back-ends for various service platforms should save development time and provide for a better time-to-market. Consolidated portals could make life easier for the end-user as well as centralizing administration for those responsible.
The Organizational Aspects
Its less complicated to deploy stand-alone technologies than it is to incorporate multifaceted technologies into homogeneous solutions.
Especially in large corporations having a complex hierarchy, it may even happen that similar solutions or services are being developed by separate workgroups or project teams. As a rule, each product manager is concentrating on his own products and each technical specialist is most concerned with solutions in his own field of expertise. Up until now that worked out pretty well.
However, as technology becomes more complex, the need for a well-defined direction and carefully planned goals increases. The constant pressure of time-to-market complicates the issue, since keeping up with the competition often cant be planned well in advance.
Organizational Approaches to Promoting Synthesis
In many ways, addressing the organizational side of the issue seems more difficult than meeting the technical challenges. Of course, as always the first step is to improve communication in general.
Raising the technical awareness and calling attention to evolving standards on decision-making levels may also help. Centralization of responsiblities for and control of the involved areas would be optimal. However, depending on the corporate structure, this may prove to be difficult.
Imagine a Service
To illustrate the concept of combining technical possibilities in a way that is optimal for the user, lets imagine a service called Home, James. Imagine you were out on the town, either your home town, or perhaps in another city. Imagine the following possible scenario: you were out and its late, its raining cats and dogs and you want nothing more than to go home.
You might be:
Standing on a street corner in your home town somewhere, you dont know where and you dont care either
In your own office in your home town
In the subsidiary of your office in another city or country
Somewhere in a strange city after a long night out and you want to go back to your hotel
The following things happen:
You take whatever end-device you happen to be using at the moment, depending on the situation
You start the service called Home, James either by clicking on a link or speaking only those words
A taxi finds you, picks you up and takes you to the desired location
The costs for the taxi get put on your bill, either on your home phone bill, your personal mobile bill, your companys phone bill, your companys mobile bill or your ISP bill
Think about it from the end-users point of view. Youre home now. Do you care how they did it ?
The majority of todays services and applications are not designed for seamless integration across diverse technologies.
There is a definite need for an architecture and standards which will allow for the integration of traditional, mobile, speech-driven and other future services, and which will allow for each service to be optimally offered to the end-users.
The technologies mentioned above are the building blocks. Its the mortar thats missing.
Position Paper WWW9 Workshop Betty Kainz, IS Technology Lab
Web and Mobility Page