Pen-based mapping - Then and now
Ronald R. Butcher Project Manager Woolpert LLP 409 East Monument Avenue, Dayton Ohio 45402-1261
Abstract
Over the last few years, a Geographic Information Systems and Information Technology group has developed a process for using pen-based ruggedized PCs for field data collection. The goal is to design simpler, more cost-effective and efficient methods of data collection and processing in the field. The process focuses on creating and verifying data during collection, not in the office. “Pen-Based Mapping—Then and Now” describes how the group, comprising staff members from Woolpert LLP, a design/engineering firm in Dayton, Ohio, used customizable off-the-shelf software products and pen-based PCs to achieve the goals. Also discussed, by case history, are the three primary application types targeted: utility management, storm water management, and facilities management. Beginning with the history of pen-based computing, this document traces the rapid progress over just the past few years. Topics include the various hardware and software technologies used to develop these systems, as well as the importance of programmability for creating a seamless process. In addition, the most recent advances pushing the pen-based world into a new generation of systems are reviewed. Introduction Significant use of pen-based mapping has only recently become possible. The steps toward developing the process follow a pattern of growth arising from needs and practices discovered during fieldwork-a common path for innovation. What is uncommon is the exponential growth that this development has experienced and will continue to show, and the implications it will continue to have on GIS and other facilities management practices. Pen-based mapping has been clearly demonstrated to effectively accomplish its basic purpose: to eliminate the duplication of effort required to generate accurate CAD records of facilities and their cataloged elements such as utilities, storm drainage, and property layout. The technology overviews and case histories that follow illustrate how pen-based mapping technology has evolved to serve these facilities management practices. History Case History : Facility Survey During the early 1990s, industries and government agencies increasingly began to demand information technology that would support graphic facilities and utilities mapping. Management executives recognized the advantages of having information that allowed a vision of a facility rather than a list of facts. Graphic layouts offer rapid analysis; database accuracy and flexibility lend the ability to explore and change a plan in an instant. So facilities management increasingly required databases that could be built rapidly and that generated site layouts, facilities and utilities representation, and survey results in graphic form. At Woolpert, Information Technology and GIS teams were beginning work on a project that would reveal the feasibility of such a process. The undertaking was a facility utilization survey of over 2.6 million square feet, comprising 237 separate facilities at Fort Belvoir. The project schedule did not allow the surveyors to collect data on paper, submit the data for electronic composition, and then recheck the results. The surveyors needed to input the data electronically to achieve the necessary level of efficiency. Using Gateway’s Handbook computer, a 486 keyboard-based “sub-notebook,” crews began to attribute database information for the survey. At just 2.2 pounds, the hardware was extremely portable-very helpful for collecting floor plan data. A technician from the group wrote a program in Clarion that supported data input screens and management. Graphics were created through storing the data in dBase 3 and using C++ to manage the graphical interface. However, the short battery life of the computer wouldn’t support the necessary software. In response, the IT team optimized the software to run on a 2- megabyte RAM card. This allowed the hard drive to be shut down, doubling battery life; by carrying a single spare battery, the surveyor was prepared for a full workday. The results of this process were mixed: The accuracy and speed desired were attained and the databases and CAD files supported the base’s user interface. But there was a general opinion that the process could be improved. For example, although facilities were fully attributed with each characteristic—from physical characteristics to content status (such as classified documentation>the system mapped just one room at a time. Floor plans could be joined to form a full picture only if every measurement was exact. In addition, using both hands to attribute the database—through keyboard entry-was awkward. The team wanted to conquer the limitations of the process, and began to source more adaptable software and more powerful hardware. The New Generation of Hardware, Software, and Programming Ca~abilitY After coming that far, the team was ready to take the process even further. Technology was rapidly catapulting hardware and software capability to new heights. CPUS were faster; hardware and memory became less expensive. New processors were being released at regular intervals and software offered more power and the flexibility of object-based operation. The pen-based computer, although in its infancy, was beginning to offer capabilities equal to tasks such as database/CAD integration. And for the first time, GPS capabilities were available in Windows. As a result, surveying and mapping software afforded rapid customization through off-the-shelf software, such as Visual Basic, C++, and Delphi. Because of these developments, a better field survey system was possible. Case History : Underm-ound Utility MatminE Field Data Collection As the team began to consider an upcoming project requiring extensive underground utility mapping, entering data via pen-based computer seemed more possible. However, the group had planned to use GPS to create accurate utility structure maps, and at that time, no pen-based laptop, or “pentop,” computer accepted GPS input. But because it was apparent that a pentop computer was capable of handling the type and magnitude of data generated by GPS, only three obstacles stood in the way. The first obstacle was devising a cable that would direct data flow from a Trimble GPS receiver to the pentop. An IT specialist studied the data and port specifications and developed the necessary cable. The next obstacle was to customize the most promising software. Finally, the process would need to be sketched, field tested, evaluated, and adjusted to achieve best practices and ensure that the new equipment and software supported them. After evaluating three software packages, the team chose a product from PenMetrics, Inc., called FieldNotes. FieldNotes is a portable GIS optimized for pentop computers using standard formats. From the first project, it was evident that the combination of GPS plus the pen-based system and PenMetrics software allowed smaller field crews to accomplish a great deal of surveying very rapidly. In the underground utility survey a crew first electromagnetically located and marked utility lines with color-coded biodegradable paint. Then, with the GPS backpack and pentop, the surveyor walked to a mark and used the computer pen to choose the “start of utility line” menu. A popup menu asked the surveyor to indicate the utility type. As the surveyor walked the line, the GPS system showed the surveyor’s position at screen center while the surrounding map changed. When the surveyor indicated the end of the utility line, a dialog box prompted for information about the line. The surveyor provided the information necessary for the system to attribute the database, and then continued. From that input, the software created an attributed vector map in the format specified by the client; in this case, the format was AutoCAD with Microsofl Access. At the end of each week, information was downloaded from surveyor pentops and merged with existing data. The process was continually evaluated and adjusted, and the flexibility of FieldNotes allowed programmers to fulfill the average software modification request from a surveyor in less than a day. Development and Resulting Productivity FieldNotes has since been customized further to better support survey practices and to perform faster than it did in that first pentop mapping project. An important capability is the option to create “smart” forms that eliminate extensive keyboard input and the associated risk of error. The user can customize the dialog box that asks for attributes so that responses are limited by the database structure. For example, when attributing an underground power line as a primary line, instead of being required to remember to attribute the field and to type the word “primary,” the surveyor simply uses the PC pen to choose “primary” on the screen. This small step, repeated daily in hundreds of input scenarios, results in time savings and a noticeably more accurate product. After the surveyor overcomes the learning curve, these advantages are apparent:
The crucial element of applying technology to the practice is exactly that: The technology must be guided by the practice, rather than the practice adapting to the technology. An important issue in the use of pen-based mapping is the customizability of the software. Every aspect of the way data is input, stored, processed, and output can be reshaped in FieldNotes. This flexibility was field tested on a citywide 47-square-mile storm water mapping project. Crew members completed an initial survey encompassing GPS structure inventory/location for storm water features. Then surveyors made a second pass to inventory inaccessible features from the first pass. A third pass completed the attribute collection. The nature of this project required building storm water processes into the GPS data collection. What the surveyor actually saw on the pentop screen was a graphic representation of the manholes and inlets. For example, by pen-touching first one manhole point on the screen and then another, the surveyor indicated that a pipe ran between the two points. The software was modified so that the resulting pipes were automatically mapped. In addition, if either of the two locations had already been attributed, the system automatically supplied size and other data about that pipe for the next junction in the mapping. As another example, if a nonterrninal drain line was indicated the system automatically showed it as a “stub”. This “connective” automation process, which is essential to building topology, eliminated much of the time and expense of recording the features found inside each manhole. Rapid Software Adjustment New issues and uses for the mapping routinely generated new modifications to the software. A programmer from the group used Visual Basic to develop many of the attribute data modifications; in addition, Visual C was used to customize input functions. This adaptability allowed information technology specialists to modify the software through simple, quick manipulation. As the storm water mapping project progressed, one issue was how to map storm water features that had been missed. For example, if a surveyor located a manhole that had not been previously identified, it was necessary to add the manhole to the map at its precise location. The team quickly incorporated into FieldNotes an icon to “place” a newly located manhole between two known points. The surveyor could touch the pen to the “place manhole” icon, and a dialog box would ask for data about the new feature. Next, the surveyor would be required to give the distance of the manhole from two known points. Then the system would automatically perform a triangulation and suggest two “triangulated” locations for the new structure. The surveyor could then choose the correct point and the mapping information would automatically be corrected. Reports of such problems encountered in the field received quick response and the surveyor typically had a solution the same day. Case History : Facility Survey Revisited The group has taken what was learned in the field and continued to adapt new off-the-shelf software for field surveys. Recently, a new opportunity arose when a previous pen-based mapping client contracted a new project with the team. Requirements of the project included creating digital floor plans for more than 60 facilities encompassing over 1.2 million square feet. The client also requested updated the mechanical and electrical facilities mapping layers for a hospital. Because the project involved not field mapping, but floor plan surveying, the software choice was Intergraph Imagineer software, which was developed to support floor plan mapping tasks. As one of the first truly object-based CAD software packages, Imagineer can be viewed as a set of CAD software tools as well as a software package. After Imagineer was optimized for pen-based hardware, surveyors set out with a package that not only was customized, but also took up only the space needed for the software elements used. Thus, with less RAM and drive consumed by software components, surveyors capitalized on more memory and faster input. Again, object-based software advanced, leading to simpler processes simpler. More importantly, the software has adapted to existing practices—fitting the solution to its purpose instead of changing or limiting practices because of the software constraints. Today Proven Hardware and Processing Solutions The strength of the combined equipment, software, and processes has been proven time and again—and not always in ideal circumstances. As one manager noted, “We have waterproof surveyors; our equipment needs to be waterproof also.” That need became reality with the advent of ruggedized, waterproof, shockproof, “surveyor-proof’ equipment. As recently as a few years ago, a drop of water on a keyboard was a catastrophe, but now a torrent is no problem. Our survey crews currently use the Walkabout “Hammerhead” Pentium pentops. Not only can they survive abuse, but they also easily accept the data stream from the GPS and support the powerful off-the-shelf software in use by crews. Advances in GPS methods have also kept pace. Recently, the team began using a centimeter-accurate procedure known as real-time kinematic GPS/GIS. It arose from the group’s mission to develop a utility GIS for the largest water and sewer district in the southeastern U.S. The project necessitated conversion of paper-system maps to a layered digital format. A strict time limitation, imposed on the client by the Environmental Protection Agency, demanded an innovative approach to rapid, high-accuracy data capture. Post-processed kinematic GPS just wasn’t fast enough. And PC data capture was critical because paper maps wouldn’t withstand the frequently damp conditions of the Southeast. A pen-based PC such as the Hammerhead could be sent out to the field, even in a severe rainstorm. Therefore, company representatives approached Trimble about developing a new real-time kinematic, or RTK—pen-based field data system that captured water and sewer structure inventory data at centimeter-level accuracies. A merger of the kinematic and pen-based technologies had been planned, but it was low on the priority list. Because the firm had a fast-track project, Trimble had only a short time to develop and test the system. Just three months later, crews had a working copy of RTK for use on pen-based computers. The resulting software, Aspen RTK, is now used on Hammerhead hardware to catalog 170,000 sewer, water supply, and storm water features, and to output the resulting GIS layers in ARC/INFO format. To date, the project has proceeded remarkably well. Future Adaptations In the near fiture, cellular links will transfer data collected in the field, which will allow near real-time data-to-image processing. Digital photography is currently being tested for incorporation in utility mapping. Planners are looking forward with realistic expectations to digital maps that offer a click-to-view process for checking details such as structural and environmental conditions. Also on the horizon are new areas of applications for pen-based techniques that may not include GIS, or mapping, finction. These include integration with barcode scanning and other processes for inventory and equipment management or security operations. But truly large facilities require databases containing many fields, and many attribution choices for those fields. These facilities will probably find that pen-based mapping is not yet a feasible solution. There is a limit to how much software, input, and information the hardware will accept. Today pen-based mapping has reached its first set of boundaries. But when the explosion of object-based software meets the next generation of processing—Intemet-based solutions, for example—the possibilities for deploying pen-based mapping will surpass today’s results and reach into realms of real-time data development once thought impossibly distant. | ||
|
|