1.2 The Internet Today
(R)Evolutions are a way of life in the computer industry. Only 20 years ago, the computer world was dominated by mainframe systems. Only a few people had access to computers, and these computers were used for calculations in large corporations. Individuals did not possess computers. The personal computer in the early eighties and the GUI in the mid-eighties changed all that, thus giving computer access to millions of people. This turned the computer into a mass-market commodity. From there it was only a small step to the Internet.
Today, more than 350 million people worldwide use the Internet. According to International Data Corp., more than a quarter of a trillion dollars' worth of business will be transacted over the Internet this year. But today's Internet is very much restricted. Although many people believe that the Internet has opened up a whole new world, it has only created a single window into this new universe. In most cases, you need a personal computer to connect to a server, and you need a web browser to browse through the World Wide Web. Companies have set up web pages to allow customers to serve themselves, reducing the load and the cost to the company. For companies going online, the benefit is clear: less direct customer interaction, higher-quality orders, and fewer problems with orders because there is no media "middleman." These factors drive down the cost for every sale and increase the profit for the company.
This is not only true for B2C web sites, but also for B2B and business-to-employee (B2E) web sites. People accessing the services need to specify exactly what they want. They need to provide a set of information and type it into the browser window. Communication is reduced from a human-to-human interaction to a human-to-computer interaction without effectively reducing the workload. The only thing that has happened is a shift of work from the business to the partner, employee, or customer. There are other advantages for the web client, of course. The company, its services, and its products have become accessible 24 hours a day, the prices have been driven down due to the market transparency, and new competitors have created an even more dynamic market place.
1.2.1 Internalized Outsourcing
Most online companies today are forced to build their entire offering virtually from scratch. Even if they buy software solutions, they have to provide all the services themselves. Amazon.com, for example, provides the service of selling books to its customers. All services required to do that, such as inventory management, distribution, billing, and web store management have been implemented and operated by Amazon.com, making their web site proprietary, massive, and costly. Although it is not part of their core business, these services need to be implemented, maintained, and operated by the online retailer. Enter pervasive computing.
A universal network will allow the next generation of online retailers to outsource these services to inventory management, billing, distribution, and web store management solution providers, which will provide these services at a lower price and a higher quality. Right now companies can do this, but they lose control over vital business functions. Pervasive computing will tightly integrate these service providers and ensure centralized control.
For the outsourcing of Internet services to become feasible, every service needs to be able to communicate with the other. The concept of service then becomes more abstract, since it is made up of a series of smaller functions. The service of billing could be further subdivided into several simpler services. One service could be the bill handling. Bills are typically printed on the retailer's printer and then sent to the customer. To reduce costs, the bill could also be printed at a local billing office, or if the customer's printers are directly connected to the Web, the bill could also be printed at the customer site. Costs could be further reduced if the bill is entered directly into the Enterprise Resource Planning (ERP) system of the company and paid automatically. But it can't stop there.
For this new paradigm to work quickly and efficiently, all levels of service need to be integrated. A new layer must be added on top of the existing Internet layers to enable services to accept other services and to connect and create new metaservices, or simply broadcast their availability to the network. Next-generation Internet startups will concentrate even more on their core business and buy the use of building blocks whenever they need them. Instead of setting up a payment server, they will rent a payment service; instead of having to buy new hardware for peak usage, they will rent network and CPU capacity from a service provider.
1.2.3 Expensive and Complex Hardware
The Internet today has the problem that a rather complex computer needs to be used to access it. Although connecting a computer to the Internet has become easier, many people are still lost if a problem occurs, because they don't understand the underlying technology of computer and networking hardware.
To make the Internet more accessible to more people, we not only need devices that are much easier to use and configure than a computer, but we also need to change the whole paradigm of how the Internet works. We are already seeing the first postcomputer Internet generation that uses more mobile phones than computers to send e-mail, chat, and search for information. For example, once the plane has landed, the mobile phone will reconnect to send out the e-mails.
Television sets from Loewe7 are Internet enabled. Mobile phones from No-kia8 are WAP enabled. But even before the WAP era it was possible to surf the Web with a mobile phone. You could either connect a mobile phone to your laptop or use the Nokia Communicator to write e-mail and surf the Web. There is also no need to buy a new television to connect to the Web. For years you have been able to buy so-called set-top boxes to add the Internet functionality to your television set. This addition allows you to browse the Web without having to know how to install a browser or update the operating system.
Using a television or a mobile phone to access the Internet is not the same as using a web browser. Later, when we discuss WAP in more detail, both from a technological and a business point of view, you will understand better why this emerging technology has not lived up to its expectations. But even if WAP were the perfect technology, it would not remove some of the elementary problems of the Web, because WAP is only a technology to access the old Internet infrastructure.
These innovations over the past years did not change things a lot, because the existing functionality of a personal computer was introduced into new devices without respecting the limitations these devices had. The concept of pervasive computing does not stop with transferring standard Internet functionality to new devices, but also allows the creation of new applications and services. Turning on a washing machine, checking prices at a gas station, locating a plane en route could become possible through the Internet. Of course, this does not mean that everyone should be allowed to access all information and services through the Internet. It is important to realize that the pervasive computing ideal of any service to any device over any network is a statement of enablement; it does not mean that every service will be made available to every type of device over every type of network.
The owner of the washing machine should be the only one to switch it on or off. The prices of petrol should be visible to everyone, but only the owner of the petrol station should be allowed to reorder petrol or change the prices. The same applies to the services and information that are provided by an airplane. Everyone should be able to check whether a certain plane is late, but nobody except the pilot should be able to fly the plane. New security models and measures are therefore necessary to implement pervasive computing technologies.
There is still plenty of room for improvement. Despite bountiful bandwidth, information is still locked up in centralized databases, with "gatekeepers" controlling access. Users must rely on the web server to perform every operation, just like the old timesharing model. Web sites are isolated islands and cannot communicate with each other on a user's behalf in any meaningful way. Today's Web does little more than simply serve up individual pages to individual userspages that mostly present HTML "pictures" of data, but not the data itself (at present, making both available is too technically demanding for most web sites). And the browser is in many respects a glorified read-only dumb terminal; you can easily browse information, but it is difficult to edit, analyze, or manipulate (i.e., all the things knowledge workers actually need to do with it). Personalization consists of redundantly entering and giving up control of your personal information to every site you visit. You have to adapt to the technology, instead of the technology adapting to you.
Another major inhibitor to creating worldwide e-business web sites is the multitude of interfaces that are often incompatible, making it impossible to share information and services across a computer, mobile phone, and car, for example. Even if the interfaces seem to be compatible, the different devices provide varying levels of data access, meaning that there is a difference between the information you receive and visualize in the car and that at home. In some cases it does make sense to present the same data in different manners, but the amount and format of data should not vary from device to device. A car should be able to receive the same data as your computer at the same detail level. On your computer, you might be presented with a map and a textual description of how to get to a certain street. In the car, the information should be presented by voice, meaning that the car computer reads the textual information to make driving safer.
Things become even worse right now if you not only read data but also use different devices for data input. Many people already struggle because they have different calendars, one on their computer, one on their mobile phone, and a hard-copy one. They need to make sure that all meetings are recorded in all three calendars. In the future, all devices will be able to synchronize themselves without manual interaction. Okay, today you can synchronize your palmtop with your laptop, but you need to install additional software and connect them via a cable. In the future, no cables and additional software will be required.
Once the technology has been put in place, personalized "information spaces" on the Internet can be created. These repositories would contain all information about a certain person, a certain process, or a certain company. The information will be most likely organized in an object-oriented database whereby objects and attributes will have security settings allowing others to view them or not. This would mean that consumers don't have to reenter information on every e-business web page, but once they decide to buy something, the site will get access to the required data. And not only web sites will have access, but other devices will be able to retrieve data whenever they need it. These devices could be owned by that particular person, such as her mobile phone, his PDA, or any other mobile device. But it could also be other devices, such as a scanner at the airport that does an iris scan to check identity.
Most web sites today focus on fancy graphics and a strong marketing message, but very few can support business and commerce transactions, and those that do, often do so inadequately. The process is not often developed by a business unit, but by the IT department, which normally has no contact with customers and partners and therefore does not know their requirements very well. Developers of web sites face another problem: the interfaces to existing systems and to partner and customer online businesses. No system today lets developers write code for a particular system and deploy it to a variety of devices. Java technology comes very close, but it means that all software components need to be rewritten in the Java programming language to be supported; many companies are not willing to spend time, money, and effort for a software porting project if they have used a particular software solution for years.
Therefore, a new paradigm and vision are neededones that address all the issues raised here and that provide the foundation for the next generation of electronic businesses. Several technologies, paradigms, and visions have been developed. The book describes them so you can choose the right one for your business case. In the end, business is all that matters.
There are more reasons to create a pervasive computing vision of the future. Consider the following example. Today, people can use mobile devices and connect to a range of devices, but in most cases special knowledge is required. Physically connecting a PDA to a computer is usually simple. The cabling is standardized, and even unexperienced users will be able to plug in the cables. The problem arises as soon as they want to transfer data from one device to the other. If they use the proposed configuration of the PDA, the transfer will work, but most people configure their computer to their needs, with tools and programs that best fit their requirements without regard to the PDA. The PDA is not prepared to download data from any calendar, it is not able to communicate with every operating system, and it is not ready to use all existing services. Today, most standard applications, such as e-mail and calendar, work, but there is no guarantee that they will work. People with more in-depth technical knowledge are able to install additional drivers and config-ure them properly to work with the system, but the technical layman will not be able to use the network.
Therefore, the existing Internet technologies, communication protocols, and interface designs are unable to handle a heterogenous network properly. The existing paradigm does not scale with increasing numbers of services and computing devices. The knowledge and time required of users today will increase dramatically as more software services and computing resources become available, and do not scale with increasing user mobility. Manual configuration costs time. If a user remains in a computing environment for only 15 minutes, he does not want to spend the first 10 minutes restoring computing contexts manually. The existing infrastructure also does not tolerate change or failure of computing resources. The services available in an environment change in the presence of mobile devices such as laptops and I/O devices. Software services can be installed or removed without users' knowledge. Proximity-based networking (such as IrDA and Bluetooth) leads to dynamically changing networks. Failures of services and networks can change the availability of computing resources. Manual configuration is simply too costly in this setting of transient services, networks, and devices.
Today's Computing Model
Today's computing model is targeted toward the individual using a single device and can therefore be characterized by the following assumptions:
Desktop computing - People typically sit in front of a desktop PCto do theirwork.
Stationary devices and software - People tend to have total control over the few devices they use, including software and hardware configuration.
Monolithic applications - Most applications are designed to interact with humans instead of with other applications.
Manual mapping - Computing tasks are mapped man-ually to applications. Users need to know which application is capable of what.
Single device computing - Users typically only use one device at a time.
Manual configuration - Users are responsible for configuring applications themselves and keep a single configuration regardless of the environment.
1.2.4 Current Restrictions
Fundamental problems prevent current technology from becoming pervasive. Today, people typically sit in front of a desktop PC to do their work. This means that there is a one-to-one relationship between the device and the human. In the future, people are more likely to use a variety of different devices that need to be configured on-the-fly as the person using it wants it to be. Even if people use more than one device today, they typically use stationary devices such as desktop workstations and a couple of mobile devices (laptops, PDAs) and they do their computing primarily on those devices. People have complete control over those devices and how they are configured. People buy software and install it on their personal devices and compute primarily with that software. If they move with their laptops from one environment to another one, they need to reconfigure the laptop to work in the new environment. Network connection, printers, and online services need to be reconfigured manually.
Today's applications are designed to interact with humans only. This means that many companies invest a lot of money in the creation of GUIs that are designed to keep the human using the software happy, but most applications are unable to communicate with other applications in a heterogenous network. CORBA and E-Speak provide the ability to offer services and information across applications and modules, but only a few applications support these paradigms todayand if they do, they support only one of them, making them available to only very few other applications.
Another problem with today's computer environment is that the user needs to know which application performs which task. Users must keep track of how to use these applications and configure them to carry out each task that they want to do. The future needs to provide a single repository where users can locate all services available to them at a certain time, at a certain location, and with a certain device. Only then is the universal network really in place.
Computers and networks are isolated silos. Communication between systems is complex and inefficient. To extend an application or technology from one computer or network to another requires additional intervention, whether it is a file conversion or a complete systems integration.
Programs are tied directly to the operating system, which is tied to specific hardware. This prevents the user from accessing any given program from any hardware devicea prerequisite for pervasiveness.
Today's users are also not accustomed to using several devices to complete a task. Everything needs to be installed on one system to be accepted by the user. In the future, the user must understand that it may be more convenient to access a service that involves several devices at once. The user should not know which device does what in this setup. It should be a transparent service, and the underlying components should be managed by the service. Single-device computing should become a paradigm of the past. A new user interface is therefore necessary to control and use the new networking service environment. This interface will also obviate the need for manual configuration that drives users crazy. Today users are responsible for configuring applications; in the future, a service will configure the application to users' current needs.