DOCUMENT IMAGING FOR BRIDGE MANAGEMENT SYSTEMS

THE SECOND CIVIL ENGINEERING AUTOMATION CONFERENCE

Sponsored by the

Metropolitan Section & the New Jersey Section,
American Society of Civil Engineers

and

Expoconsul International

Jacob Javits Center
New York, New York
November 19, 1991

Avanti C. Schroff¹ and Sanjiv Nathwani²

ABSTRACT:

This paper introduces document imaging technology and explores its use in bridge management systems. Document imaging is considered a vital first step to true multi-media systems. A detailed solution for BIN folder management is presented as a specific application of document imaging in bridge management.

INTRODUCTION:

A bill for the first major overhaul of the federal transportation system in 35 years is currently in review at the House of Representatives. The bill has already been approved by the Senate. The impact of the bill is clear: Emphasis will shift from new construction to maintenance for existing structures. Reportedly some $13 billion will be allocated over the next five years exclusively for bridge improvements. Further, spending will be decentralized to allow greater freedom for states in using Federal transportation money. New York State has responded be more than doubling its spending on bridge maintenance and repair for fiscal year 1991.

A particular emphasis has been placed on developing bridge management systems to oversee and coordinate this increase in activity. It is an opportunity for the bridge management community to rethink and reshape their current level of computer automation. Strategic use of state-of-the-art technology should result in enhanced productivity and cost efficiency.

Bearing this in mind, this paper introduces document imaging as an innovative technology that could revolutionize information flow within the industry. Its strength lies in its ability to convert documents such as hand written notes, engineering drawings, blueprints and even color photographs quickly to electronic media. This would bring the information on-line for communication over high speed networks. Document imaging could effectively replace overloaded, widely distributed and geographically dispersed physical document libraries with all their attendant problems and limitations. The realization of a paperless office might very well be at hand.

To illustrate the power of this technology, the bridge inspection process for New York State Department of Transportation has been selected as a case study. A summary of the current procedure is followed by a detailed description of how a document imaging network can be implemented for the process. It is hoped that by analyzing the benefits of imaging for a specific sector, the possible impact of this technology on the industry as a whole can be envisioned.

DOCUMENT IMAGING: TECHNOLOGY OVERVIEW

Document Imaging, also known as Document Image Processing (DIP), Electronic Information Management (EIM), or simply as Imaging, is the computerized management of documents which were originally in the "hard-copy" (paper, photos, etc.) form.

In concept, Document Imaging is simple. Pictures of documents are input, stored, accessed, displayed, printed, and transmitted like all other types of computerized information. However, because images of documents are pictures, not letters or numbers, and because these images use a lot of computer storage when compared to other types of data, special programs and equipment are needed to handle them.

The simplest variety of Document Imaging system is an image database, where a user can store and retrieve not just facts and figures, as in an ordinary database, but document images as well. An example of this would be a database of properties for sale; the user would be able to view pictures of the properties, images of inspection reports and deeds, as well as descriptive text information and numeric prices.

Ordinary computers may be used for imaging. PC’s, Apple Macintoshes, and UNIX workstations can be employed as the desktop workstations, while a variety of systems can act as servers in a networked configuration.

Special hardware is required for entering, displaying, and printing images. A digital scanner is used for document input. The scanner works much like a fax machine or a digital photocopier; in fact, a fax machine has a digital scanner as one of its components. The scanner is connected directly to a computer, which controls it. Because documents are stored as pictures, high-resolution displays are needed on the viewing workstations for image to be legible. Such displays typically have resolutions of at least 72 dots per inch (dpi). Standard laser printers can be used for printing black-and-white images. However, a special printer interface may be needed to get acceptable printing speeds, due to the large size of image files. Specialized color printers would be required to print the color images.

Also, due to the storage requirements of images, special storage devices are often mandated. Though images can be stored on normal magnetic disks, just like other types of data, it is often advisable to use optical disks due to their high capacity (600 megabytes to 10 gigabytes) and removability. Though magnetic floppy disks are removable, their storage capacity is small (no more than 20 megabytes), and conventional hard disks, though much faster than optical disks, are not removable, thus limiting total storage. Until recently optical disks could not be erased or changed once written, but rewritable optical disk technology is now economically competitively with write-once (WORM) systems.

Scanned images are usually compressed before storage and are decompressed for display or printing. Compressed images use less disk space and can be transmitted over networks more quickly. Black and white images are usually compressed using CCITT Group III or IV compression schemes, resulting in substantial reductions in data storage requirements. Color images may be moderately compressed using Lempel-Ziv Welch (LZW) compression, which perfectly preserves the original picture, or the Joint Photographic Experts Group (JPEG) scheme can be employed for high compression ratios with minor image degradation.

Standard local-area-networks (LANs) can be used with Document Imaging systems. However, because images are large data objects, low-speed wide area-networking (WAN) may be prohibitively slow. Faster WAN technology, such as T1 or Switched Multimegabit Digital Service (SMDS), may be more appropriate.

DOCUMENT IMAGING TODAY:

The technology to construct Document Imaging systems has been available for well over a decade. In this respect, Document Imaging is a well understood and established technology. However, it is only recently that the technology has come to the forefront of mainstream computing. Several factors have contributed to this emergence:

• As noted earlier, managing images on a computer requires specialized hardware. Until recently, items such as digital scanners, high-resolution monitors and high-speed laser printers were prohibitively expensive, and only the largest installations could justify the expenditure. The meteoric growth in the graphics art industry in recent years has resulted in mass production of these specialized components, and consequently affordable prices.
   
• The recent revolution in optical disk technology has finally presented the computer industry with a practical and affordable medium of storage for large data types. This technology, more than any other recent innovation, has been responsible for making Document Imaging a practical alternative in data management today.
   
• Escalation of the battle for the fastest micro-processor has resulted in affordable workstations that provide adequate response for manipulating image data in real time.
   
• The introduction of algorithms for image compression and decompression have greatly improved network access times and reduced storage for image data.
   
• Last, but not the least, legislative bodies have established precedents for the use of optical disks and imaging in document archival. A bill is currently on the floor in the Senate to precisely define the manner in which this technology may be used for storage of legal documents. This acceptance has effectively opened the floodgates, and many conservative institutions such as banks are now embracing imaging wholeheartedly.

Document imaging reportedly will be the area of most prodigious growth within the computer industry for the next five years. Studies have projected that in ten years some 95% of business documents will be in image form. Some $1.2 billion will be spent on imaging systems this year and that spending is projected to grow to some $12.5 billion in 1995. The technology promises to change the basic infrastructure of industrial organization. The advent of both word processors and database management tools have simply added to the tide of paper. Contrastingly, Document Imaging promises to replace the paper deluge at least to the point that paper documents become discardable items. It offers an opportunity to convert all non-editable documents to electronic form rapidly. This should help unleash the true potential of network communications; until now its power has been held in check by the seemingly impossible task of having all documents, regardless of their source, keyed into computers by data entry personnel. Further, documents that were hereto incapable of being converted to electronic data, such as photographs and drawings, can now be integrated in context with normal text and numeric information.

Whether or not Document Imaging manages to find the holy grail of business computing, "the paperless office", it promises to radically alter the state of computer automation to a more efficient and comprehensive state. It also paves the way for full multi-media systems.

DOCUMENT MANAGEMENT IN BRIDGE INSPECTION:

The volume and type of information generated as a part of the General Bridge Inspection process for NYSDOT and the communication of information between the Regions, the State Central Authority and a consulting engineering office, seem to best exemplify the sort of problems that can be alleviated by the introduction of Document Imaging.

The most significant documentation to consider are the BIN folders. These are largely created and replenished by the bridge inspection process. Other processes that contribute to the information in these folders are repair, rehabilitation and design records. These folders contain all pertinent information for each bridge. The information is comprised of location maps, inspection forms, remark sheets, flag reports, color photographs, inventory data, load rating data, special emphasis sections, sketches, engineering drawings, and other miscellaneous information.

Typically, each bridge is inspected once every two years as a minimum, either by the state personnel in the region or by a consulting engineering firm. After each bridge is inspected, the information is compiled and added to the contents of the BIN folder. The BIN folders generally are held by the consulting engineer until the inspection of a particular bridge is completed. As the inspections are completed, the folders are returned to the region headquarters and particular documents in the BIN folders are handled over to NYSDOT central authority in Albany. Here, pertinent information is keyed into several mainframe databases for tracking. This procedure is modified only in the event of a structurally deficient bridge, in which case, the bridge is inspected every year until the deficiencies are addressed.

There are several inherent problems with this method of document management. Some of the problems are common to all processes that result in paper files:

• Documents are occasionally misfiled and thus inaccessible. Recovery of the misplaced document may entail several hours or days of manual searching. This is compounded in situations where documents may reside in more than one location.
   
• Paper media deteriorates over time and through continued use. BIN folders are constantly referred to by many different departments since they house the most recent and complete information about the bridges.
   
• Paper media is single access only, i.e. only one person may effectively review a given paper document at a time. In the case of BIN folders there are numerous instances where many people would like to view the same information at the same time but in diverse locations.
   
• Paper documents are easily damaged or lost while being used. This is particularly true of BIN folders, since they are often carried to the site for bridge inspection.
   
• Paper documents have poor access, i.e. paper documents are bound to the geographic location where they are stored. Anyone requiring a document must either go physically to the storage location or request that someone else do so to make a copy for delivery.
   
• There is generally a significant lag between the creation of a document and its delivery to its point of storage. For instance, the lag time for use of BIN folders can vary from 4 months to 2 years depending on the scheduled inspection dates for bridges within a two year cycle.
   
• Physical document libraries are extremely space inefficient. Filing paper documents physically requires an enormous amount of space when compared to electronic document filing.
   
• It is extremely cumbersome to obtain specific information from physical files. For example, an arbitrary request for all inspection rating sheets would require collecting these sheets from each BIN folder.

Other problems unique to this specific case are:

• In the case of structurally hazardous conditions, documents need to be communicated to all parties concerned in as short a time as possible, along with color photographs.
   
• While all text and numeric information is capable of being keyed into one of several database management systems both at the region and central authority levels, it is impossible to commit color photographs and engineering drawings to electronic format using conventional means. This means that currently, only a fragmented view of overall bridge information can be brought on-line, thus severely reducing the usefulness of any network communications efforts.
   
• Completed BIN folders are transported back to the region. This transportation takes place at least once for every two year cycle and amounts to a significant effort and cost.
   
• The other significant documentation is produced by the contract management process. Every two years the job of inspecting the bridges in a region is re-implemented. Apart from the contract document itself, there is a continuous stream of correspondence to contend with during the two years, both for reporting and accounting purposes. These documents are prone to the same management problems as those described above for BIN folders.

DOCUMENT IMAGING FOR BIN FOLDER MANAGEMENT:

BIN folder management is an ideal application for Document Imaging. The technology can bring test and numeric information that is contained in the folders on-line immediately. The fact that text documents would be treated as uneditable images is not a drawback in this case; the process is structured so that information need only be added or replaced to the BIN folders. In addition, photographs and drawings can also be brought on-line using the same mechanism and be capable of being viewed in context with the other information.

There are about 20,000 bridges in the eleven regions of New York State. Each region may contain anywhere from 800 to 2,000 bridges. Generally, there is at least one BIN folder for each bridge. Assuming an average of about 100 pages per BIN folder, the total storage required for these 20,000 bridges would be approximately 200 gigabytes.

Any new computer system of this magnitude is best implemented in stages. There are several reasons for this:

• The expenditure can be distributed over time to make the process more affordable. Requisitioning funds to build a state wide network in one stretch would be difficult.
   
• Phased implementation will guarantee that personnel training and workflow changes keep pace with the functionality provided by the system in each phase. It would be inefficient to build the complete system and wait until the industry adjusts to it over time to realize its full potential.
   
• Phased implementation allows input to the process from users at every stage. A one step installation assumes that user needs have been completely thought out and represented at the onset. This is almost never the case.
   
• A phased approach will allow new technology to be incorporated as it becomes available. This is a particular concern, since the computer industry is evolving almost day to day. There is a cost efficiency benefit as well; trends clearly suggest that specialized hardware required is getting progressively less expensive.

Examining the phased implementation of a Document Imaging network for BIN folder management illustrates incrementally the benefits that can be derived. For the purposes of this paper, total development is outlined in three phases. Several more might be necessitated in a real world implementation.

PHASE 1: AUTOMATE BIN FOLDER MANAGEMENT AT THE REGIONAL LEVEL

The goal for this phase would be to completely automate the management of BIN folders for each region and to provide networked access to information to as many users as necessary within regional offices.

Since several points of access are required, a client server configuration is deemed optimal. Thus, the system would be constructed with a central data server connected to several workstations over a LAN. The primary tasks of the server would be to manage the central database generated by the Document Imaging application and to respond to database requests from the workstations. The workstations would be used to scan information to the central database and view/print information locally.

BIN folders would be scanned into the system as soon as they are received by the region. Similarly, new or updated information would be added to the system as soon as it becomes available. This would ensure that all information on-line is current and the system can be used as an authoritative source for all BIN folder information.

All personnel within a region would access the information by viewing the documents on-line. Printed copies would only be generated for revisions, and distribution to consulting engineers, field teams and NYSDOT central authority. Fax transmission capabilities could be added to further facilitate off-site communication.

The benefits to each region after the completion of this phase would be:

• The application would be designed with extensive error checking for data entry. This would greatly reduce the probability of misfiling. Accessing information on-line does not involve temporarily removing the documents from the database as does getting a document from a filing cabinet. This would reduce misfiling opportunities. Further, in the event of a misfiling, an on-line search is considerably faster than a search through a physical library.
   
• Image information would be stored on optical disks. Optical media shows no deterioration with repeated use, and is extremely durable. Paper documents would become discardable and additional copies could easily be generated over laser printers on the network. This would effectively remove the need to track distribution of BIN folder documents.
   
• Since the network would provide simultaneous access to BIN folder documents to all required points of access, several people within the regional office could work with the information at the same time.
   
• Adequate backup procedures would ensure that information once entered into the system is never lost. This would ensure that no information is accidentally lost or destroyed while in use.
   
• Condensing a paper filing system into an optical disk library should free up rooms of filing cabinets at regional offices for other uses.
   
• BIN folder documents can be cross-indexed to offer extremely sophisticated access to the information. This would greatly improve the turn around time for researching existing information.

PHASE 2: CONSTRUCT WAN TO ALLOW CENTRAL ACCESS TO REGIONAL DATABASES

This phase would allow NYSDOT central authority to communicate with regional offices via high speed communication lines. Central authority would be required to implement a document imaging system of its own. The database for the central authority would be a subset of the regional databases, with only relevant information being transferred over the network to reside at the central offices. Procedures would have to be devised to ensure that the central data subset is kept current.

Alternatively, central authority could opt to simply maintain a remote node with access to all regional networks. This would facilitate on-line querying of regional databases with view/print capabilities, without the expenditure for a network to manage an in-house database.

The benefits derived from this stage are as follows:

• Central authority would have on-line access to BIN folder information. Regions would no longer have to mail documents to central authority as BIN folders are modified during subsequent inspection cycles of rehabilitation contracts.
   
• There would be no lag between information available to central authority and information at the regions.

PHASE 3: ADD CONSULTING ENGINEERS TO THE NYSDOT NETWORK

This phase would allow all organizations that are involved in the bridge inspection process for NYSDOT to have access to the BIN folder information on-line. Again, the firms could opt to implement a full document imaging network at their premises, or simply maintain view/print/scan workstations. In either case, they would be able to access BIN folder document over the network.

Consulting Engineering firms contracting for bridge inspection work would have the ability to modify the information on the regional database directly, as new information is produced by their bridge inspection teams. This would ensure that most current information is always available at the regional level and at the central authority.

The benefits offered by implementing this phase would be:

• Information would no longer have to be mailed to and from consulting engineers and the regional offices. This should result in considerable cost savings.
   
• BIN folder duplication would no longer be required, since all parties concerned would have on-line access to the information. If paper copies are still required, they can be printed directly from the regional database. This should result in additional cost savings.
   
• There would be no time lag in communicating inspection information to the region and the central authority. New information would be added to the regional database as soon as it is compiled at the firm performing the inspection. This might be particularly helpful in situations where joint decisions need to be made regarding the structural integrity of a specific bridge.

Implementation of the total system would close the information loop for the bridge inspection process. Inspection documents would become available throughout NYSDOT almost as soon as they are created. Similarly, all other modifications and revisions would be noted instantaneously by the entire network. The burden of transporting information would be reduced to field teams, who would carry only pertinent information on-site. Paper copies would become redundant and expendable, resulting in lower costs for space and for personnel to manage the paper libraries.

The overall impact would be to dramatically improve the quality of information and rate of information flow. Such streamlining should result in an overall efficiency that would save the time and money in bridge inspection cycles.

DOCUMENT IMAGING AND BRIDGE MANAGEMENT SYSTEMS

Considerable effort is being expended both by federal and state organizations in developing comprehensive Bridge Management Systems (BMS). Emphasis has been placed largely on developing applications that will refine the priority scheduling for bridge maintenance and rehabilitation. Reportedly some 50% of all bridges nationally are structurally deficient or functionally obsolete and in dire need of rehabilitation. With limited funds available for this gargantuan task, it is easy to see why such immense care is required in allocating limited funds to the most needed structures first.

However, BMS has no significant impact on the assimilation and dissemination of the core information for decision making. This information is collected in the field and in non-electronic form. Despite the eventual implementation of BMS, this information flow will continue to be plagued by the deficiencies of a paper intensive process. Document Imaging offers a unique opportunity to alleviate this problem and contribute to the overall efficiency of the bridge management process. The benefits are well illustrated in the study of the bridge inspection process provided above. It should be easy to extrapolate these benefits to all other paper intensive sectors of the industry.

In addition, Document Imaging may provide the missing third dimension in on-line analysis of structures for scheduling considerations. By the very fact that visual analysis by field teams is required to assess the condition of a structure, it would be of tremendous benefit to augment text and numeric data accumulated in BMS with photographic evidence.

CONCLUSION:

United States, perhaps more than any other nation, is almost embarrassingly dependent on automobile transportation and thus the highway network that sprawls over the country. Cars are a veritable symbol of a fiercely independent American lifestyle, and the move to mass transportation that is being encouraged is not likely to meet with early success. Some 500,000 bridges help connect this maze of roads and they are in severe need of rehabilitation and replacement.

The bridge management community faces the unenviable task of keeping these bridges operational and thus America’s highway network alive. It is extremely important that no stone be left unturned in the struggle for efficiency in this process. Computers are perhaps the most powerful tools for efficiency in today’s society, and keeping up with the state of the art in computing may offer the edge that the bridge management community needs to stay on top of the problems. This paper offers Document Imaging as a powerful technology to wield in the industry’s endeavors to battle the tide of paper and other information.

¹Avanti C. Shroff, Senior Vice President, Iffland Kavanagh Waterbury, P.C.

²Sanjiv Nathwani, President, Trilon, Inc.

TRILON, INC. 528 Winfield Avenue Upper Darby, PA 19082-2122 (215) 790-6265 voice (215) 790-6231 fax Info@Trilon.com http://www.Trilon.com