| Back to Top | White Papers | Keith & Faye's Home Page | Photo Albums |
This article was originally created in 1989 and was updated in 1992. Although many of the predictions have now come true five years later, it is still instructive to review what has happened in those five years. Particular insights may be gained from the clarifications of the terminology of object-oriented systems. Comments for 1997 are mode in italics
There are several parallel developments that were brought about by the PC revolution that are leading to the complete integration of processing for data, voice and image. The cost of storage is dropping rapidly, and the widespread availability of optical disks is driving costs down further. Relatively low cost image scanners, graphics displays, voice capture cards, voice response cards, and FAX attachment cards are making the PC a practical vehicle for handling a wide variety of the data "objects".
The term "objects" refers to data records, software programs, scanned images, libraries, spreadsheets, printer images and voice messages used in cooperative processing systems with host-managed intelligent servers and workstations. The availability of inexpensive OS/2, Wintel and UNIX processors makes handling of large objects feasible.
Object is used to distinguish each item from the traditional data record. The computer handles, stores and distributes objects without regards to their content. The content is only meaningful to the user at the destination when using a "transducer" like the PC to make the object useful by presenting the contents. The implications of these trends to the industry are covered in the attached document, and some recommendations are presented under the following sections:
The 1980's saw numerous organizations taking advantage of their private telecommunications networks by combining the bandwidth needed for various data networks with that of their internal voice networks. This had two key advantages: economies of scale associated with internally-managed high-bandwidth networks, and the ability to dynamically allocate bandwidth among different applications based on the time of day. Unlike the telephone companies who have to accommodate instantaneous demands for service, enterprises have the advantage of being able to plan for peak demand times in their networks based on their understanding of the business volume flows during the day and week.
This integration involves specialized controllers at each end of a high capacity "pipe", such as a T1 link. Portions of the pipe are allocated to the specific applications to handle their planned load. The PBX (Private Branch eXchange or voice switch) does this whenever it "connects" two telephones for a voice conversation. It depends on unused moments to get more traffic on a given pipe. Similarly, the on-line network can be given more capacity at noon for consumer demand or during morning and afternoon peaks for office use. Large capacity is available during low periods to transmit large objects such as FAXes, spreadsheets, and programs.
Newer applications of PC technology include full motion video delivered to the desktop. In this application, video feeds from cable, satellite and local media are distributed via LAN to the client workstation. The image is digitized and compressed. Only changes in the video image are sent over the LAN, greatly reducing the amount of data transmission needed.
The PBX stores the digitized voice objects on disk for retrieval and distribution using their VoiceMail applications. It acts as an object management system for these specialized objects.
Imagine your e-Mail In-Basket showing item #5 as "Voice" rather than Reply, for example. When you select #5, the voice message plays through your PC speaker (or handset if privacy is needed). You can forward or store it, while Reply will start the microphone to record your reply which you send along with the original. You might reply with an eMail note just as easily. The eMail could be created using your favorite PC editor or word processor. The distribution lists, copy lists, nicknames and file folders would all be common to both voice and data objects.
Compound notes and documents are a natural extension. You might get a note highlighting an article from a local newspaper which is attached in a scanned image format (FAX). You can send it along to a team-mate attaching a voice instruction or comment to them. All three components are available from the menu screen when you review it. All the elements of this technology are operational today in I.S. departments and in various development groups.
The standard phone call has almost been forgotten. Have you ever considered what a treat it is to call someone and actually get them answering their phone the first time? Usually there is some secretarial screening or tag as the norm. The original system of connection was simply a set of relays connecting the two telephones together momentarily with copper wires. There were no recorders at either end, and no translation took place anywhere in the system. The system dynamically allocated part of its bandwidth (2 copper wires) to achieve the communication.
Today the phone call can be handled by allocating a portion of a digital communications channel with enough bandwidth to provide sub-second response to a voice object entered into the network by the desk-top telephone - you say a sentence (for 5 to 8 seconds), and it is digitized by your desk-top phone continuously. The digital data network gets it to the other end, and it is translated back to an analogue reproduction in under a second. As long as there are no copper wires, all the bandwidth can be re-allocated while there is a pause between the two parties, to take a breath or to think, for example.
During this pause, the same channel could be used to transmit a 10K object that was a compressed FAX page, without the voice users noticing. Both the voice and the FAX could be captured on the same PC technology. The controller of the T1 link must be capable of giving voice objects highest priority, then data transactions, then images, for example.
The potential is even more dramatic when you consider that much live and expensive long distance bandwidth is wasted today dictating to a remote voice recorder. Considerable savings are possible simply by dictating to your local recorder and scheduling distribution in the next fifteen minutes to the remote VoiceMail system. The system now can utilize low usage moments to send the voice object. Broadcasts to remote parts of an organization become much easier and more personal than the text messages being used so much today. Because the system manages everything, the cost is much lower than the physical creation and distribution of videotapes, for example, or even printed copies.
Current technology also makes the presentation of full-motion video to the desktop feasible. Using compression and intelligent scanning that only sends the modified part of the image (such as the movements of the head and lips by a speaker, for example), normal PC connections can transport such messages for playing on demand by the recipient. Simple adapters running under OS/2 or Windows can present the reconstructed motion video in a window on the desktop PC.
Such technology introduces the opportunity for just-in-time education at the desktop, by calling up a centrally stored and managed videodisk, or by linking via satellite to such organizations as the "video university" currently operational in the United States. Regular communication by the executives become much easier and ad hoc, while the need for costly training rooms is substantially reduced.
What will take time is the maturing of software and standards needed to manage the intermix of data object types. By the time that software is available, the hardware will be sufficiently inexpensive to justify immediate implementation. If everything were in place today, it might take up to five years to address the planning, standards, capital needed, and organizational implications for rolling out a fully-integrated processing system.
What can you do to position your organization to exploit the technology more quickly? There are a number of simple applications available today which can be justified on a small scale to allow you to install the base technology for future integration:
On any particular item, the proliferation of vendors and choices is a mixed blessing. While it makes for ample selection and low price, the proliferation of management responsibilites in most organizations is leading to piecemeal implementations and duplication of equipment and expertise. It is also robbing the end users of the absolute joy that can accompany a system-managed integrated support system,
Central management and common technical support are critical to making the integration happen. This does not imply centralization. With the technology available today, the expertise can be close to the "front lines" were it is most effective, while the supporting technology is invested in only once. In fact, the technical support community will be one of the biggest beneficiaries of Processing Integration because they tend to be the most advanced users of the technology today.
Although many of these dependencies will get implemented as part of the associated applications, it would be better if they were justified on their own. In this way, the justification may not have to cross organizational boundaries. They can proceed as separate projects removing them from your critical path to Processing Integration:
No vendor has the answers. Many are motivated to continue the proliferation which requires duplication of costly technology investments. Unfortunately, the real cost of all the current duplication is the cost to operate, maintain and support all the duplicate equipment and services.
Many organization have manual systems for filing, copying, and distributing hard copy. They also maintain people at switchboards and message centres. The real irony is that 99% of the information in the hard copy originated in digitized form, either from a PC application, a computer report or a digitized voice!
Processing Integration is the Nineties opportunity that mirrors the technology opportunity from the Seventies achieved by American Airlines and in the Eighties by FedEx. Yet it is open to all industries and government. Enterprises that achieve Processing Integration and successfully connect their clients to this rich infrastructure will be around in the next decade to share the secret of their success...
He has had experience in managing quality investment processes, transforming services companies, and is working with a small but emerging group of associates specializing in breakthrough approaches to training, management, business engineering, and leadership.
Keith is currently a Senior Partner with Yorktown Technologies Incorporated and is responsible for the practice of Implementing Executive Strategy. Yorktown Technologies is a company that specializes in breakthrough business re-engineering, as well as object-oriented development technologies like SmallTalk, and artificial intelligence systems. They have developed proprietary models for the IBM Neural Network Utility which make that technology practical for many more businesses now.
Copyright the authors and/or Keith Cowan