Undertaking a new document management program can be a daunting and complex task without extensive research, experience and education. With the help of an industry certified professional such specialized tasks as system design, hardware requirements and information capture can be streamlined and simplified.
When designing a new system everything from storage requirements, database configuration, bandwidth and security need to be addressed. Calculating storage requirements can depend upon the types of documents being stored within the system. An average letter size page scanned in black & white at 300DPI will require 50KB of storage space. This means that a department scanning 2,000 pages per day with an expected 5% increase in volume after the first year would require over 130GB of storage space after five years. With the relatively low cost of storage media this may not seem like much over a five year period, but this is only one department and migration and back-up requirements still need to be addressed. Database configuration also plays a key role in determining the requirements of a document management system. Will your selected document management system integrate into your existing database server or will it exhaust your system resources? For instance, some document management products will actually store files and images as "blobs" within the database, where others simply use call-outs within the database to retrieve the images from a dedicated storage pool. The industry preferred method, however, is to use call-outs to reduce the burden on your database and improve performance. Meaning you have a greater chance of utilizing existing system resources without further investment.
Selecting the right hardware and capture software can lead into a sea of manufacturers, brochures and specifications. Asking the question, which is the right choice for your business? One of the most important factors to consider when selecting a document scanner is daily volumes. This can be based on the peak or an average daily amount. This number should be sufficiently large to prevent an ongoing backlog. Virtually every brochure will list the maximum images per minute the scanner will perform at with all factors being optimum. These numbers never represent a true working environment. For instance, simple math would tell us in order to scan 50,000 pages in an eight hour shift; we would need a scanner capable of 105 pages per minute. However, this doesn't take into account higher resolutions, image clean up features, orientation correction, operator fatigue and loading/unloading time. Even though every scanner is different a commonly used calculation for determining scanner throughput is:
Rated Scanner Speed x Average Throughput (70%) x Number of Working Minutes Per Hour (50mins) x Hours Per Shift = Actual Daily Throughput
Using this calculation a scanner rated at 90 pages per minute will actually only achieve 25,200 pages per shift, not 43,200.
Capturing index values and other critical data is another hurdle in a document management roll-out and can be performed in a variety of different fashions, including:
- Barcoding
- Manual data entry
- Forms processing
- Full text OCR
- Zone OCR
- Handwriting recognition
- Automatic data extraction
How do you find the right balance of accuracy and efficiency? What methods produce excellent results automatically? And what methods can result in documents lost permanently?
Many methods of capture can reduce the time spent manually entering/correcting information. Barcodes, for instance, tend to have a superb accuracy rate. If your document management software is able to recognize an accurate barcode on the document it will extract the proper information. If the software is unable to recognize the barcode the field will be left blank and easily identifiable during quality control.
Barcoding matched with database extraction allows you to populate all index values based on one highly accurate and unique barcode field. This same database extraction method twinned with an OCR process can actually lead to very small, yet very critical, percentage of records that are mis-indexed. Imagine if you were to OCR a health card number from a patient record, and the OCR results we're in accurate. When the system queries the database for the remaining index values it could potentially populate the fields with another patient's information losing a vital patient record.
The above examples indicate why professional document management companies employ industry certified specialists as part of their sales and technical support teams. They come prepared to better assist in making informed decisions critical to the success of your document management solution and operating efficiency.
Find out more about ECM solutions by visiting our website
You can also follow us on Twitter www.twitter.com/MESHDS