By Anthony Macciola at Mon, 01/03/2011 – 14:29
Reposted from the AIIM Capture Community Blog

It’s not uncommon for organizations to want to lower their overall total cost of ownership for IT infrastructure. Capture owners are no different. The use of Intelligent Document Recognition (IDR) technology has delivered significant cost savings in relation to minimizing and/or eliminating human touch and involvement throughout the capture process. These cost savings are usually realized during the runtime execution of the capture system.

The other area that can be leveraged relative to lowering the overall total cost of ownership of a capture system is in the setup, configuration and ongoing management of the system.

Historically, capture systems are set up and configured as standalone, independent systems. This often will result in the creation of redundant batch, document and field taxonomies and business rules that typically exist within existing IT infrastructure and systems.

If you take a step back look at the big picture, why couldn’t the capture system leverage and exploit these existing taxonomies and rules whenever possible, or use ground truth data, taxonomies and rules whenever available? It could and should.

If you’re wondering what capture systems are likely to look like in the future, they’ll likely be extensions or integrated components of existing IT infrastructure or line of business (LOB) applications and will be managed and executed under the purview of a larger LOB application versus being a standalone entity.

Capture’s roots are based in archival, namely capturing paper documents (usually after a business process has executed) for the purpose of archiving the paper documents in a repository of some sort. As capture transitions beyond scan-to-archive and becomes instrumental in process automation, the historic independent capture system mentality will be forced to evolve into a more transaction centric model that is controlled and integrated within an organization’s application framework. In other words, capture will evolve to become an integrated component of larger LOB applications.

This is a subtle but material evolution for the capture market and will allow capture to expand to a wider range of applications and process automation use cases.

When you combine this evolution with some of the other innovations and trends we’ve talked about over the past couple of months, it should become evident that capture is going through a renaissance and is poised for a new era of growth and expansion.

By Anthony Macciola at Fri, 10/15/2010 – 17:24
Reposted from the AIIM Capture Community Blog

So what’s the first thing that comes to mind when you think about business process automation (BPA)? I’ll bet that it’s not a multi function peripheral (MFP)! Why do you think that is?  When you think about it, many business processes are document driven and are initiated in the field as an ad-hoc transaction. Historically, people have used fax as a way to initiate a transaction, but we all know the image quality limitations that come along with fax.

However, there is another, somewhat obvious element to consider.   Many of today’s organizations have network attached MFPs deployed in the field, so why not allow a knowledge worker to initiate a business processes by capturing the document that drives the process? In other terms, why not leverage your fleet of enterprise MFPs to be the front office interface for your field based knowledge workers and your back office BPA infrastructure. It’s the perfect marriage and an optimal and effective use of your overall IT infrastructure (doing more with less, so to speak).

That’s why any enterprise capture platform must have an MFP component that provides the ability to centrally discover, manage and monitor a distributed fleet of MFPs.

However, there is one slight catch.  Most MFPs don’t produce the same quality of image output that users have come to expect from production scanners. Or, at a minimum, most MFPs lack the ability to capture mixed document sets and product reliable and predictable image quality. For that reason, many enterprise customers will typically use production scanners rather than leverage their existing MFP fleet.

Traditionally, the image processing/perfection provided in most (if not all) MFPs is geared toward providing aesthetically pleasing print output as opposed to process-ready output (process-ready being a document that is conducive to downstream document classification and metadata extraction, both of which are key precursors to any business process automation framework). When you look at the most common use of MFPs in the capture world, it’s traditionally been based around collaboration. This includes scan to fax and scan to email, where the knowledge worker receiving the document is likely to print it as well. In this case, everything works fine. However, things start to get a bit sketchy when one uses the MFP for archiving the content or initiating a business process.

This discussion sounds very similar to the mobile phone capture subject we discussed earlier.   In concept, the potential problems share some similarities.  For example, how does one simply and reliably produce process-ready documents?  This is an issue that Kofax VirtualReScan (VRS) has solved in the production scanner market. Turns out the same problem needs to be addressed in the MFP market (as stated by a couple of leading industry analysts). By solving this issue, an MFP could be a viable alternative to a production scanner, especially in a distributed, transaction oriented business process automation environment.

As I mentioned when we discussed mobile devices, the MFP platforms haven’t evolved to the point where we can get access to raw CCD output at the level we need and currently get from production scanners, but we’re seeing significant forward progress and interest as various MFP providers realize the upside and opportunity to differentiate themselves and effectively compete in the BPA market. It wouldn’t surprise me if we saw new generations of MFPs brought to market that had VRS embedded and were able to produce process-ready documents straight out of the device itself.

Such an evolution would be a win-win all around as MFP vendors would now be able to position their devices as distributed business process kiosks (much higher value than a shared output device) and customers would be able to consolidate their overall IT infrastructure, saving money while increasing overall efficiency.

This is definitely something to watch and plan for as the business process automation and management markets become more pervasive and dominate the overall buying behaviors of global organizations.

As a leader in the document capture market, we’re committed to making this happen and evolving the world of document capture to the next level.

By Anthony Macciola at Fri, 09/10/2010 – 13:56
Reposted from the AIIM Capture Community Blog

How do you know what kinds of documents you’re capturing? How do you figure out what types of documents are entering your organization? These are the kinds of questions that need to be asked in today’s document-driven organizations in order to maximize productivity, increase efficiency and reduce costs.

Most people today are still using the “old school” method of using document separator sheets with patch codes and/or barcodes on them. Although very effective, this often comes at a cost; materials (paper and ink) and labor relative to manually reviewing each physical piece of paper and determining the appropriate place to insert the preprinted separator sheet.

For those of you still using this method, it’s time to embrace the 21st century and think about how you can best drive cost out of your capture operation by eliminating the need for separator sheets.

The compelling replacement for separator sheets is document classification. There are two distinct technologies used for classifying documents.

• Image-based classification:Looks at the geometry of the image, based on image layout and patterns; can be trained by showing the classifier samples and labeling them appropriately.
• Text-based classification:Looks at the content of the document and is based on text patterns; can be trained by showing the classifier samples and labeling them appropriately.

Classification is also a critical component of document separation. Document separation technology leverages classification techniques to determine what it’s looking at and then applies additional logic to ascertain document boundaries and logical groupings. Together, classification and document separation are a cost-effective alternative to document separator sheets.

In addition, document classification can have a significant impact on other aspects of your overall capture operation.

• Mailroom Operations: classification can be used as a tool to discern what is being captured and where it should be distributed.
• Capture for archival purposes: classification can be used to determine the appropriate folder for the captured content.
• Capture for the purpose of driving a business process: classification can be used to ascertain what has been captured and determine the appropriate workflow to initiate.

Effective classification offerings require little to no prior setup. If you’re using a solution that requires scripting or dictionary (keyword) creation and/or maintenance, you should re-think what you’re doing. Competitive classification offerings require nothing more than showing the system a variety of samples (tens of samples vs. hundreds or thousands) and labeling them. From there, the system can and should do the rest.

If you’re managing a capture operation you should be focused on how you can minimize the amount of human interaction/intervention required to maintain operations, be it scan preparation, scanning, quality control, indexing or otherwise. Classification is one of many tools that you can utilize to increase the overall efficiency of your day-to-day operations.