April 1997
15 Imaging Gurus Predict the Future of Our Industry
Discover where the industry is headed. Prepare yourself for "living documents," "intelligent doors," "200-ipm scanners" and imaging
Imaging as a Sales Tool
Harry Newton, Publisher,
Imaging Magazine
We live for our customers. Nothing is more important. A picture is worth a thousand words. That's why we sell with images. TV is the most powerful mass sales tool ever.
To date, imaging has been a back office thing. It focused on fixing horrendous paper problems. It's done great. There have been huge customer benefits. Utilities can now solve customer bill problems in minutes, not weeks.
I want imaging to be pro-active. I want it to thump the customer. "Buy my product. Do it now."
Take the TV marketing mentality and apply it to imaging. Two trends make this thinking imperative:
1. There's been an explosion of products and services. Most consumers now have a hard time distinguishing between one overnight package delivery service and another, between one bank's service and another, between one company's insurance policy and another's.
2. What makes the difference between products -- why I choose one over another -- is a blend of subtle differences of two flavors. First are complex technical and financial differences. The second is what I call "The Surround Stuff." Information that surrounds the product and how easy it is to get that information. "How much do I have in the bank?" "What's the best place to sit on your plane?" "How do I fix my washing machine?" "How do I get to your store?" "What's my hotel room going to look like?"
These and the thousands of questions our customers ask are best answered with images. But they're not images we throw at them on TV. They are images tailored to one particular customer.
The next phase in imaging is the creation of complex imaging systems to deliver those one-of-a-kind images instantly to the customer.
The elements of this next imaging revolution are already here:
1. The Internet. Ubiquitous, cheap, easy-to-use and multi-platform.
2. The ubiquitous Web browser. Runs on all operating systems on all computers. Any computer with a Web browser (and all have them) can get on the Internet (private or public) and get picture information off the Web and private intranets.
3. The PC fax board. You can now put 96 "fax machines" inside one PC and blast or tailor faxes from the broadest variety of databases -- visual and non-visual. The software -- especially the simple computer telephony applications generator -- has improved dramatically in ease of use.
4. The lowly fax machine. It used to be slow, boring and expensive. No more. There are fax machines running at 64,000 bps -- over six times as fast today's most common machine -- the 9600 bps machine. You should see these newer, fast machines fly.
5. The call center as a customer imaging hub. Agents at call centers pump out specialized faxes by hitting F7 on their keyboard. "Want to fix that washing machine? Here's an exploded view of your washing machine. Try that. Call me back in 15 minutes if that doesn't work. We'll try something else."
6. The Fax as a Binary File. Every PC running Windows 95 is a fax machine that will accept faxes and deliver them. If the faxes are sent by another PC running Win95, the "faxes" will arrive in an editable file with words and pictures. The recipient can change them and send them back. This is a major plus for marketing.
7. Image enabling of databases. True story: A hotel chain finds if it puts images only in front of its agents (but not its customers), it sells more upgraded rooms.
The agents get bubbly about how beautiful the better rooms are. The excitement in their voice convinces the customer on the other end of the phone line.
Show me what I'm interested in, I'll buy. Show me how to fix it now, I'll be your customer for life.
Imaging works.
Imaging's Past and Future
John Mancini, president,
Association for Information and Image Management
International
What's changed in the world of imaging technology during the last decade? Just about everything. In the early days of imaging, most systems were either small implementations, with less than 10 users, or one-of-kind, custom-developed, large-scale solutions. The larger implementations took years to develop, were proprietary and fraught with the land mines inherent in emerging technologies. The early imaging systems tended to support high-volume, transaction-oriented systems.
We've learned a lot since these early implementations. Not just about how the technology works, but about what needs can be meet with imaging and what other technologies we need to integrate to produce business solutions.
Four primary forces have molded imaging over the past decade and they'll continue to shape and reshape the industry into the next century:
LANs have emerged as a core business technology. Many organizations have migrated to client/server-based architecture. LAN-based architectures have brought down the cost of imaging systems. The hardware is no longer as specialized as it was, and the LANs are already in place, so software is now the most significant investment. This trend, which will continue in the years to come, has made the entry into small pilot imaging systems easier to swallow.
It's still hard to move to enterprise-wide applications. Many vendors and users expected a ubiquitous image retrieval mechanism to emerge and they assumed the systems in place would expand enterprise-wide. Unfortunately, this hasn't happened. The nature of the applications and the organizations' business processes make enterprise-wide imaging different from departmental transaction-oriented systems.
Imaging isn't a standalone technology. The transformation of the imaging industry has occurred on another level. Many organizations have discovered that imaging is just one application in a larger technology portfolio that can deliver complete document management solutions.
The VAR channel is ascendant. Because users must integrate various image components and other technologies to create an overall solution, partnerships and alliances are widespread. This has had an effect on the selling strategies of systems integrators and value-added resellers. The channel now provides a strong conduit for document management solution sales.
Does imaging have more than an application future? Is there more to life than imaging at the desktop? The answer is yes.
The ability to recognize and search for a document -- not just with a text-based search but by color, shape and other visual features -- takes imaging to another dimension. Compound documents will provide an overall framework for managing new-age data -- electronic text, image, graphics, video, multimedia -- challenging the dominance of record-oriented information.
Imaging Will Enable Everything
Dr. Charles E. Stenard,
advanced image recognition solutions manager,
Lucent Technologies
Over the past five years, we've heard claims about imaging's enormous potential. Now we're seeing successes with real-world implementations.
Today's successes are the result of intellectual capital built over the last half-century. Pioneering organizations have been working on imaging for years. As early as the mid-1970s, many groups believed it could be applied in narrow vertical niches like medical imaging. These early adopters received benefits great enough to outweigh the upfront investment.
Just recently, though, drastically reduced costs and improvements in software have moved imaging far beyond what we thought possible only a few years ago. Some systems now can read imported fax images and use workflow to route them throughout the enterprise.
What's on the imaging horizon? I see the technology evolving to include accepting electronic and HTML forms -- and other kinds of messages. Once you start viewing imaging as an enabling technology, you see it playing a substantial role helping organizations improve key business functions like transaction processing. Paper-intensive industries like banking and healthcare are already benefitting from implementations.
In the not-too-distant future, imaging applications will include voice and video, and applets will be embedded into the messages. And why stop there? Why can't we design a system that lets users receive messages in their choice of media, be it hard copy or text-to-speech?
I see a day when this will all be possible. It's simply a matter of thinking beyond current applications. We've moved so far in my 20 years in imaging that I no longer believe that anything is unattainable. I just wonder how quickly we can achieve it.
Internet Imaging
to Explode
Samuel DenHartog,
president, CyberLink Labs
In 1997, imaging, document management and COLD will continue to converge. The result: document-based information management systems and central information repositories.
For the new information to be accessible within the corporation and to the broader world accessible by global communication, the information management infrastructure must be based on open technologies, including TCP/IP, HTML and Java. These exciting intranet/Internet technologies, along with the new network computers, will provide improved access to information on documents at a much lower cost. We'll see a lot of growth in the number of companies combining imaging systems with intranet/Internet technologies over the next year.
Terabytes in Your Pocket
Lee H. Elizer, founder,
DataThink
Storage is an interesting challenge today, but it will be the new frontier in the next decade. Imaging has redefined online mass storage. As a result, applications expand as fast as the price of storage hardware plummets.
Remember when word processing was a new thing? When kilobytes of storage held a "bunch" of letters in the late '60s and early '70s? When spreadsheets, the "hot" new application, were the most useful way you could use 100 KB of storage? When PCs became inexpensive with a 10 MB hard disk and a floppy disk held less than 1 meg? When CAD/CAM became the engineering application that generated humungous storage requirements (megabytes) in the late '80s and early '90s? When multimedia arrived de facto in newly purchased PCs in the ý90s? Today, Internet applications and their use abound. Everyone's storing images, e-mail and downloaded files. Storage needs on today's PC easily exceed two gigabytes.
But think about the storage explosion that will occur during the next decade, containing exciting new applications and yet-to-be-determined "killer apps" on the market:
E-mail with audio -- should arrive in the mid-to-late '90s.
Voice-activated PCs (with artificial intelligence) sporting interesting new applications in word processing, spreadsheets, e-mail, imaging -- late '90s to early millennium for popular use.
Cyber cafe "drop-in" centers for Internet access -- late '90s. The buzz at these cafes: How does one store downloaded information and take it "to go?" Smart cards with hundreds of megabytes of storage? Portable small tape cartridges? Removable hard disk drives?
Digital cameras on each business PC and remote "teleoffices" will become a reality. Videoconferencing from the desktop and e-mail with real-time graphics will appear in the late '90s to early millennium.
E-mail with images captured at the desktop will result in unprecedented capacity requirements from this and other new Internet/intranet applications by the year 2000. We can also look forward to dial-up movies on the PC.
Imagine the day when we all carry a few gigs in our pockets: text, images, voice-prints, fingerprints. There's no question that extremely high capacity, interactive storage at the desktop will arrive by 2005. The technology for this type of storage is now being developed by forward-looking companies who will integrate the storage for the next millennium.
Imaging Will Transform Management Styles
Geoffrey James, author
Business Wisdom of the
Electronic Elite
Improvements in imaging technology will change the social structure of corporations all over the world, throughout the next decade.
More information will be stored inside document databases, eliminating much of the interoffice junk mail that now clogs our desks. Massive rows of filing cabinets will disappear, because the most current version of any document is available online. As a result, an increasing number of workers won't have to report to work each morning. They'll merely telecommute to where information is stored. Or access the data from their homes.
At the same time, improvements in mobile computing, multimedia and direct video will make it easy to coordinate projects using workflow and groupware. It will become less necessary to have people working in close quarters merely to coordinate projects.
These changes in technology will force managers to abandon today's "command-and-control" style of management. A result-based model, where workers are paid for projects accomplished rather than for warming seats, will become more common.
This will force the social structure of the corporation to become more decentralized -- a meritocracy rather than a bureaucracy. Corporations that embrace the new technology and the new style of management that it demands will grow and prosper. Corporations that adhere to old management will discover that the combination of industrial-age thinking and information-age technology is entirely toxic. Old-style organizations with empowered bureaucrats and control-obsessed micromanagers, will falter, decline and ultimately disappear.
Document Management is Dead, Long Live....What?
Michael C. Row,
Vice President, Business
Development, EZPower
In the scramble to jump on the internet bandwagon, many in the electronic document management industry overlook a far more ominous transition -- the death of document management as we know it. An extreme statement? Possibly, given that EDM has barely emerged from infancy, according to some experts. Plus, the business issues addressed by document management (information access, security, and control) are increasingly critical, even more so with the growth of the web.
Still, the value-added of stand-alone document management systems (DMS) is declining. The new model will integrate EDM functionality with line of business applications databases and distribute this new functionality across the enterprise. This shift, from library-oriented EDM to database-oriented EDM, is already changing the industry.
Library-oriented EDM/DMS emerged to tie together all an organization's document-based information: imaging of paper documents, COLD, word processing files, spreadsheets and other electronic documents. Products from vendors such as FileNet (including Saros & Watermark), Documentum, PCDocs, and EZPower provide a unified and consistent way to organize, access, and control document-based information without users having to be concerned with the technical format or physical location of the information. The primary metaphor of these systems has been a library or repository of documents.
Now, the traditional value added for stand-alone DMS is increasingly commoditized, part of the operating system, or both. Databases, groupware, and internet/-intranet services (like Netscape's catalog server) are aggressively encroaching on areas previously the province of DMS.
Capture (scanning) and storage functions were originally part of DMS. Now they are highly competitive component markets in their own right. Traditional DMS vendors will have their work cut out competing against the focus and expertise of Kofax, Cornerstone, Cheyenne and Optisys.
Plus, databases are being extended with object technology to permit flexible storage management, while operating systems are extending storage management functionality -- the NT 5.0 distributed file system will provide most of the functionality now found in top of the line storage products, including jukebox management and HSM functionality.
Basic library functionality of indexing and security is under heavy attack from the groupware market, the internet market, and operating system development. Microsoft Exchange, Lotus Notes and Novell Groupwise all contain the basic indexing and security capability of a stand-alone DMS, plus many more critical capabilities. Netscape and Microsoft's catalog servers also incorporate advanced indexing and security, and the upcoming NT 5.0 file system will have indexing built in.
The problem of viewing documents with many different technical formats has largely been solved. Omni-viewers (such as Inso's QuickView) are pervasive and increasingly cheap and/or provided with the operating system. Moreover, document conversion is also highly advanced, to the point where Netscape and Microsoft are distributing conversion filters with their web servers.
If basic library services are ubiquitous and nearly free, what is the future of EDM? The next frontier in EDM is database-orientated EDM, leveraging existing application databases to provide flexibility and control in document organization and retrieval.
In this model, documents will be treated as part of the organization's information model. EDM functionality will be tightly integrated with key line-of-business applications such as human resources, sales-force management, and customer-service. Yet documents will still be accessible throughout the enterprise in a controlled and consistent fashion.
But What A
Data Type
Peter Hamilton, president and CEO, LAVA Systems
Reports about the "death of images" have been greatly exaggerated. True, the image viewer, as a marketable, discrete application, has a limited future. With free viewers on the Web and ActiveX controls being bundled with Internet Explorer, imaging services are becoming part of the standard desktop. What this really means is that imaging, far from dying, is now a pervasive technology, with scanned documents widely accepted as "just another data type."
More and more large organizations now use imaging as an efficient way to access information traditionally held on paper. The small office/home office market is discovering imaging through the use of fax modem software and inexpensive scanners. Consequently, the number of people encountering scanned documents at the desktop is growing exponentially. Vendors are even starting to talk in terms of the "casual" image user.
What's most important to these users is the content and value inherent in the document. They don't care if it's an image or another data type.
Value can best be added to an image document by relating it to other information sources.
Users want to access and work with documents, of whatever format, within the context of their core business systems -- perhaps even from the user interface of an existing application.
To support this need, many vendors need to rethink their product architecture. They must move towards a distributed computing model, where the necessary elements of imaging, document management and workflow technology are object-based and can activate on demand, wherever they are needed.
This will let organizations provide users with a single point of access to all business information through an interface that offers the methods and manipulation tools appropriate to the data type, but which appears seamless and intuitive.
In a world where document images are everywhere and any computer user can access and view images for free, the focus of development must shift from the problems of image access to new opportunities for image use.
The future of imaging lies not just in the ability to access images, but in the incorporation of that ability into a larger, more comprehensive information access architecture.
Smart Objects Blend Imaging with CAD
Chris Dukes,
imaging product manager,
Softdesk
In the CAD (computer assisted design) area, object technology holds the future of imaging. Industry-leading CAD vendors have begun to make this trend possible by including technology that lets developers add their own objects into the CAD environment.
For instance, Softdesk's Autodesk included the ARX (AutoCAD Runtime eXtension) development environment, which enables the creation of objects, in its latest version of AutoCAD software.
An object, in this case, is a CAD entity with intelligence. For example, a door in a CAD drawing is typically a collection of lines, but a door object "knows" what its appropriate behavior is. (For starters, it knows that it cannot exist in space -- it must be within a wall.)
Applications developers can narrow the range of possible behaviors to the set of appropriate behaviors. What does this mean for imaging? With object technology, even scanned images can become objects with intelligence.
Object technology lets you work with black-and-white, grayscale and color raster files as easily as you can work with CAD entities. It enables raster images to actually become part of the CAD model and of any electronic document.
Previously, raster images (which are composed of dots) were difficult to integrate with vector drawings (which are CAD drawings and models defined by mathematical coordinates).
But object technology lets you work with raster and vector data as one object. This means you can edit both raster and vector simultaneously.
Web Imaging Is
Your Future
David Skok,
chairman and ceo,
Silverstream
The World Wide Web has introduced three new and very important standards to the world of electronic documents and electronic messages:
1. A universal electronic document format, HTML, which everyone can read.
2. Universal e-mail addresses.
3. Web sites with Web addresses.
All three eliminate paper: HTML simplifying the exchange of electronic documents; e-mail by replacing fax and mail; Web sites by eliminating physical duplication and distribution of documents.
Obviously, there are huge benefits to keeping information in electronic form, using e-mail and the Web for storage, distribution, sharing, and management. The result: paper will be a secondary form for documents, used only where it is more convenient to read. Many paper transactions -- loan applications, insurance forms, etc. -- will become electronic.
The elimination of paper threatens imaging and document management. The Internet and intranets allow corporations to build a single universal architecture for distributing, sharing and processing information. You don't need to turn paper into electronic documents if the document is electronic to begin with. You don't need special document management software if you have an intranet.
For document management vendors, this means serious new competition from Microsoft, Netscape, Oracle, Informix and IBM/Lotus. These players are drawn to the enormous opportunity presented by intranets. They are shipping products that include features that used to be the purview only of document management systems.
Corporations will find current document management systems offer only a small piece of the pie. They want the complete suites of servers now offered by Microsoft and Netscape, with document management tightly integrated.
Some additional predictions:
1. Imaging, document management, groupware and web servers are converging into a single information repository, to be stored in a new generation of universal databases from Oracle, Informix and others.
2. The Web browser will be the universal viewer for all information types. You can use it to view HTML, PDF, JPEG, GIF, Word files, PowerPoint, sound, video, VRML, etc. It's only a matter of time before TIFF image viewing is included.
3. Successful vendors will be those who move to thin Web clients first. This means offering imaging/workflow/document management clients that are a mixture of dynamic HTML and Java.
The Document: It's Alive!
Joe Forgione, VP of business
multimedia products,
Lotus
The 1990s, characterized by the phenomenal growth of the Internet, have ushered in the most significant evolutionary wave for document media. Today, documents are not only electronic vehicles for communication and collaboration, they also offer "interactive" interfaces to information.
A document today can contain a narrative created in a word processor, financial analysis created in a spreadsheet and links to databases or Web pages. Like Frankenstein's monster, the document has become an electrified, living creature.
Acceptance of diverse data streams -- e-mail, bitmaps, fax, HTML, voice, alphanumeric pages, video, telephony, paper -- has spawned demand for a personal solution to content management within networks and I-nets. The future lies in liberating users from the Herculean task of managing this essentially unmanageable mass.
Perhaps users need technology that automates a document's life cycle. Maybe we'll see documents that arrive ready-made with preprogrammed business rules, or stored logic, letting documents determine their own review cycles, classification, routing, migration and eventual self-archival.
These so-called "intelligent agents" (user-defined attributes or rules) could move documents through a business process without end-user involvement, changing along the way to reflect new rules for each consecutive step. Set attributes could define the next series of reviewers, the classification of the document and the file cabinet in which the documents should reside. And since documents reach a point where they are rarely accessed, rules could define the transfer of old documents from the file server onto lower-cost storage media.
Pipe dream? Not really. Soon, this will be the status quo, as we begin to embrace even loftier forms of communication, such as 3-D Web sites, Web broadcasting and video streaming.
Capturing Images
Faster Than Ever
Serge Moisan, president, Dunord Technologies
Scanners as we know them, slow and fast, will become faster. The trend toward affordable 200-plus-image-per-minute (ipm) production scanners will continue.
We should see scanners capable of processing more than 500 images a minute within the next five years.
Slower-speed scanners will advance too. They will continue to offer more features and color.
Super high-speed scanners bring two new challenges:
Because it is impossible for operators to inspect images so fast on the fly, image quality control will have to improve. Standards will emerge for document and forms applications to communicate quality control criteria. This will let scanners and scanning applications perform automated image quality assessment without any human intervention.
Applications won't be able to process documents as fast as they are captured -- even with upcoming generations of PCs.
As a result, more sophisticated pipelined scanning subsystems will come to market, along the lines of InputAccel from Cornerstone Imaging (San Jose, CA 408-435-8900).
While the capture side of our industry is gearing up for this kind of evolution, users will demand better overall throughput, more accuracy, fewer rejects and better fidelity. Black-and-white images, too limited in content, will be replaced by grayscale images and subsequently full color. This will challenge integrators to learn how to deal with much larger images.
For years, hard disk manufacturers have battled about how to connect drives to PCs. We're beginning now to see that neither IDE or SCSI will win. The same will happen for image capture.
Scanner manufacturers will want simpler and faster interfaces to the PC. The current battle between SCSI and video interfaces will see no winner. Instead, fast serial protocols will become the interface of choice.
The Future of Text Retrieval
Bert Rankin, Vice President of Marketing, MindWorks
Though text retrieval technology will evolve in an incremental and linear direction, one thing will not change. Vendors will still have to answer the following questions if they are going to meet the needs of their customers:
1. How do we manage all the information we deal with daily?
2. How do we retrieve that one specific piece of information we need?
3. How do we best use that information, once retrieved?
Physically managing the information should become considerably easier, particularly when we're talking about paper-based documents.
Image-aware applications, coupled with low-cost scanners, more powerful processors, and larger storage devices should give just about everyone the flexibility to scan and store documents without having to worry about adjusting or choosing things like OCR settings, file names or compression algorithms.
As more desktop systems become image-enabled, the transition from static document to searchable archive should be almost complete. But look out for document overload. Each new source of information, from digital cameras to live news feeds over the internet to near-realtime print media, will drive the relentless cycle of more, faster, better.
The technical challenge: to help people find all their information.
Most developers recognize full-text document retrieval as the only practical means for accessing information, even for photographs and video. With full-text retrieval, the emphasis shifts from the arbitrary (document title, creation date, application used) to the specific (what information do I require).
Simply type in a search-word and get a list of possible matches. Users of the Web are familiar with this form of search and retrieval. They also know its limitations. A complex indexing system requires more time to index a document, limiting throughput. Simple systems speed throughput but make it harder to find the exact document you need.
And, as the sheer volume of available information increases, the indexing systems of today simply cannot meet the requirements. Referred to as inverted indexing, these define the absolute location of a document based on its attributes stored in a central index. Great for archives, not so good for the dynamic documents which are evolving out of document management, internet and office suite technologies.
Compound, OLE-enabled, interactive, distributed and portable documents generate new requirements because they constantly change shape, change location, change content. The indexing and retrieval system best suited to this new world of documents is actually one which binds itself to the document itself (including Portable Document Index -- PDI -- that's offered by MindWorks).
In its current form this type of index allows users to share documents over LANs and the Internet, to distribute them via CD-ROM, or simply to shuffle them between laptop and desktop systems.
Future versions of the PDI could perform a variety of additional functions, based on the environment and intelligence embedded in it. Just as Web documents can today perform "housekeeping" duties using Java applets, or gather information using Bots, so too could documents equipped with modified PDI capabilities.
The technology at this point becomes the glue, binding the various document components together as it registers itself with search engines, moves from media to media, updates itself, or retires itself from circulation into a data warehouse.
Forms Processing Enters the Mainstream
Reynolds Bish,
president, TextWare
Forms processing -- or automated data entry -- has been with us for a long time. Ten years ago, it required large-scale, proprietary scanners with hardware based, in-line recognition technology. These systems cost more than $1 million, so few were sold.
In the five years since Imaging Magazine first rolled off the presses, the cost of forms processing systems has declined by an order of magnitude.
Scanning, image enhancement and recognition technologies are better, less expensive and easier to use than ever. Client-server platforms with open architecture are more powerful and affordable than previous systems. As a result, the market is much broader than it ever was before.
Still, most forms processing systems take only centralized, batch-oriented approaches that prevent the adoption of scanning, enhancement, and recognition technologies in a wider range of environments and applications
This will change during the next five years as forms processing merges with mainstream electronic document management systems.
Both types of systems capture images and data. In document management, the emphasis is on the image. Erroneous index data can be tolerated through the use of multiple indices or fuzzy search algorithms. With f