The technology of electronic document management systems (EDMS) is a relatively new one, at least for small and moderate sized businesses. They have been in use on very large mainframe systems for a long time, but it is only fairly recently that EDM has become a viable, useful and increasingly much-needed resource for any size of computer network.
An EDMS is basically a computer database system designed for keeping track of documents, changes to them, related documents, and possibly to control the access to documents and assist or even enforce established procedures for workflow or QA. An EDMS can track all kinds of documents whether in computer data form, books, paper, audio and video recordings, microfilm, or anything else. For computer data, an EDMS can provide detailed and automated access to the data. For other types of documents, it can index where they are stored. A major function of an EDMS for all types of data is to enable items to be found quickly from a wide range of search criteria.
The term 'document' is used in this context to mean absolutely any form of information storage. In the case of computer data, an EDMS can provide viewing, printing and editing facilities for a wide range of data types such as text, spreadsheets, graphics and databases. Some systems have special facilities for design data, usually including support for large format scans of old drawings as well as proprietary CAD data such as DWG files and similar.
Sure signs that EDM is needed in a business are:
* People sometimes do modification work on data files that are not the latest version.
* There is confusion about data that is in progress, being checked, currently issued and superseded.
* It is not always abundantly clear what is the current version and where it is.
* Conflicting versions can be found in various places.
* Superseded versions cannot be found when needed for historical reference.
* There is uncertainty about what other documents are related to the one being altered.
* Records of changes, checking, approval and issues are inadequate.
* Information cannot easily be found unless the drawing numbers or exact titles are known.
* Excessive volumes of paper prints are made for very short-term internal reference.
* There are multiple data storage locations.
The whats and wherefores
For some offices, the main value of EDM is being able to find things and keep track of superseded versions. Others adopt it more for the ability to control the authorisation of changes; that is, QA procedures. Now that storage space is so costly and the need to provide for anti-disaster archiving of data is more recognised, many firms are scanning all their old drawings and microfilms just so that they can be compactly stored on multiple CDs and indexed by the EDM system for fast access.
An EDMS can store indexes to all documents that are related in any way to each document, so that whenever any document is accessed, the list of related documents is shown. This can be extended to itemise those related documents that have design dependencies on the data in the viewed document, and may be affected or made invalid by changes to the current document. In the case of drawing data, it can itemise all linked or referenced external drawing files.
For design data such as CAD, EDM can provide much more widely accessible viewing of CAD data, making it available to anyone with access permission without needing full CAD software. This can be done either by Windows graphic viewing applications installed on all computers, or even more flexibly, by implementing HTML-based CAD viewing systems so that each computer only needs a web browser, which is probably a standard facility.
Furthermore, such viewing applications and web-display systems can provide a facility for authorised users to mark-up the drawings for alterations. This is called 'redlining'. The redline data is stored in separate files so that the CAD data is not changed, but is intimately linked to the CAD data so that it is always displayed with it. Where this facility is used, it usually becomes the standard method of defining changes and tracking their implementation. The redline data is un-linked when the changes to the CAD data have been completed and approved. The check-in/check-out facilities in the EDMS ensure that all authorised CAD operators can have access but only one may be carrying out changes at any time. Any other users during that period can view or print the data that was checked-in for changing, but cannot alter it, and are warned that it is being altered.
These procedures are especially valuable where design, review, checking and authorisation may be spread over widely separated locations. An EDMS such as DataViewer can operate with all facilities over a wide-area network linked by dedicated lines, ASDL, ISN, satellite links or other means.
EDM systems range widely in size and facilities. The large systems, many intended for mainframe use, are extremely costly and require extensive specialised customisation for each situation and a long implementation period. Until relatively recently, those were the normal form of EDMS, and it was only extremely large organisations that could consider their use. More recently several very small and simple systems have been introduced, and several mid-sized ones. The smallest ones are often just file indexing systems. DataViewer has all the facilities mentioned in previous paragraphs. It can be implemented very quickly and easily by local staff with minimal disruption to on-going work. The basic tailoring of display forms to the users needs and procedures can be easily set up. More elaborate customisation can be done but is often not needed, and can be implemented gradually over time.
Some EDM systems provide control of data access by complex systems of encryption of the database and hiding of the documents in the database. This type of solution is referred to as a 'vault'. The benefit is the documents are invisible or unrecognisable other than via the EDMS. With the demise of several EDMS solutions it is now seen as undesirable complication and introduces a risk of major problems in the event of equipment or supplier failure. The latest EDMS security solution takes advantage of the now available 128-bit encryption. The solution works by encrypting the files, which prevents unauthorised access to the files. The benefit is the ability to keep your existing file names and directory structure while having complete control of access.
The data and bases
Any EDMS is based on a database management system (DBMS) and needs to maintain a single database for all its indexes, relationships and operations. The very small systems often use an internal DBMS of simplistic form. Others use one of the several well-established network database systems. Some require a particular make of DBMS, while others can operate with any of the common industry standard systems. DataViewer can operate with a wide range of common database systems and can therefore make use of a company's main system if desired. It can use a shared-data-file system such as Microsoft Access for small networks and data volumes, or a Client/Server SQL system for any size of situation. Client/Server systems run continuously on the data server, and provide reliable high performance operation regardless of the number of users and the volume of data. DataViewer can work with Oracle, SQL/Server or Borland InterBase Client/Server systems. InterBase is included with DataViewer and is popular when the user does not already have a Client/Server DBMS. This adaptability to various data systems makes DataViewer easy to implement in any situation. It also allows for easy transition of an initially small system to a large Client/Server system if the company grows, without the disruption of changing to a different EDM product and re-loading all the data.
It is important that a system of suitable initial scale, future scalability and capability is implemented at the outset. Several companies have regrettably demonstrated the cost of choosing a system with highly proprietary mechanisms that cause severe and costly problems if that vendor's system ceases to be marketed. Here is a checklist to help avoid those problems:
A factor in achieving a cost-effective EDM implementation is to begin with only those facilities that will, in the particular business, provide major advantages. However, it is important that the system should be capable of being expanded to add the facilities that initially were judged to be 'nice to have', rather than 'must have'. Setting up any data system is inevitably a significant task. It is important not to risk having to start over again later with a more powerful system.
No matter how many great advantages an electronic data management system may offer to a company considering adopting the technology, a major factor in the decision will be not only the operational benefits to be gained from the outlay, but also actual cost benefits. It might be feasible to justify an on-going extra cost for the work benefits obtained, especially in some critical fields of work, but usually the new system must repay its cost in terms of money saved in day to day operations.
If you refer back to the check-list at the beginning of this article - the circumstances that indicate the need for an EDM system - it will be clear that if those issues exist, overhead costs will be considerably higher than they need be. It is of course hard to quantify the cost of delays and confusion caused by difficulty in finding data or working on the wrong version of data. In the case where work proceeds for some time on data that is later found to have been the wrong version, the wasted time and money can be considerable, or even lead to penalties.
However, experience with companies that have implemented EDM systems of one make or another, indicates clearly that, quite apart from operational and convenience advantages, the actual cost benefits from an EDM system quickly pay for the system's acquisition and implementation. From then on, the on-going cost-savings and operational advantages provide a return on investment.
It is possible to work out some of the cost factors involved in data access, and hence to estimate the saving likely from using an EDMS. To do this, a company needs to obtain some measure of the time taken to perform data access tasks with the current manual system. This will differ considerably for various companies. Generally, an EDMS will almost eliminate the time currently taken to find the right data.
If there are eight people on say $20 an hour and each consume even only one hour a week finding data (which is pretty good going), then that amounts to 8 x 20 x 1,67 = $267,20 per week, or $13,894 a year! (The figure '1,67' is the magic number generally considered to represent the indirect cost of an employee.) This figure does not take into account the cost of the computer and software licences sitting idle during this search time. Even at this rate, an EDM of the cost of DataViewer, pays for itself within a year. Check your own office situation - it might surprise you!
Clearly, the larger and more complex the existing network, the more time is likely to be needed to find the right data. But even if only small time-savings are indicated for a small simple network, they will become much greater as the network grows. System implementation will be much quicker and cheaper when the network is small. Adding data as work proceeds is never a problem - it is loading the existing data initially that is laborious. The sooner that is started the better!
DataViewer can be implemented initially as a very simple system on a small network, or on a large scale system initially, and can gradually grow as the company's activities expand, and can have facilities added to it. It addresses all EDM needs from very small to very large without needing complex and specialised set-up. It meets all the choice factors listed previously.
* The EDMS should use industry standard mechanisms, such as its database system.
* The data should be stored using standard computer file systems.
* The EDMS should be able to be expanded indefinitely in terms of data volume and number of users.
* The EDMS should be able to operate across several network servers and long distance links.
* The EDMS should be useable initially with minimal configuration done by local staff - ie, not need costly up-front consulting services.
* The EDMS should be programmable to add new facilities.
* The EDMS must be able to handle all the types of data that the company may need, including viewing and printing facilities for them.
* For design firms, the EDMS must be able to handle CAD data.
* Changes to the configuration should be able to be implemented while the system is in operation without needing down-time, and able to be done gradually in the light of experience, by local staff.
* If starting with a small system, the EDMS should not initially need excessively large and complex mechanisms, eg, its database, but should be able to move to more powerful mechanisms if the system grows, without needing re-implementation.
* The opportunity should exist for adding facilities that may not be wanted initially, eg, QA workflow control.