I am always amazed that a typical day includes between 10 and 20 packages, with considerably more heading out during our peak season. All contain different types of geospatial data that are ready for use by our clients for a multitude of applications.
But it’s not just the numbers that continue to surprise me; I am also somewhat amazed at how our delivery system has evolved over the last several years.
Ten years ago, Mylar or paper plots were a required deliverable for virtually all of our projects. Today, I'm not sure I could even find a roll of Mylar for one of our plotters. What’s more, I can’t remember seeing a shipping tube in the last 18 months that would accommodate the plots in an outgoing delivery stack. Delivery tubes in our office have been replaced with DVDs and, more frequently today, portable hard drives. In fact, more than 350 portable hard drives were shipped from our office last year, and those totaled more than 200 terabytes of storage. Also, understand that these numbers were compiled for just one of our office locations, not across the entire enterprise.
The Data ExplosionSo what has changed in this profession to result in such a significant shift in the way we deliver the results of our professional efforts?
For one, detailed planimetric line maps have been replaced, for the most part, with digital orthophotos. Instead of interpreting two lines that represent the edges of pavement for a roadway or the limits of a sidewalk, our clients want to view the actual road or sidewalk in an image so that they can make inferences well beyond simply the physical location of the road. In addition to location, they want to see important qualities like pavement type, surface conditions and physical appearance.
A similar argument could be made regarding buildings in a mapping project. An image contains significantly more information than a closed polygon that only represents the building footprint. In addition to the added information, digital orthophotos can have a significant advantage over planimetric maps in terms of considerably reduced project costs if planimetric feature compilation is reduced in lieu of an orthophoto deliverable. Horizontal accuracy is not sacrificed in digital orthophotos; accuracies here should be similar to planimetric maps created from the same controlled imagery.
But these improvements come at a cost in terms of data storage. Where a typical planimetric map for a 5,000-foot by 5,000-foot tile might be measured as a few megabytes, an uncompressed color digital orthophoto of the same area at a 6-inch pixel resolution would be around 300 megabytes. At 335 megabytes per square mile (a 5,000-foot by 5,000-foot tile contains 0.90 square miles), it doesn’t take long to accumulate a lot of digital data on an orthophoto project.
LiDAR is also a contributing factor in the data explosion. This incredibly capable technology has resulted in a number of large mapping projects for our office over the last several years. One of the primary reasons is cost. Elevation models developed from LiDAR are created at a small fraction of the cost of models developed from stereo compilation of controlled aerial imagery. But the LiDAR sensor is indiscriminate and places the same density of points on the ground throughout the project area regardless of whether the ground is flat, rolling or mountainous.
This differs from a traditional mapping approach where the number of ground points placed by a trained compiler would be appropriate for both the amount of undulation in the ground and the required vertical accuracies. Compilers place significantly fewer points in flat or more “predictable” areas and considerably more in areas with substantial and unpredictable elevation change. This task is labor intensive, but it results in an efficient digital elevation model that, when combined with breaklines, produces much smaller files than a comparable LiDAR elevation model. For example, one of our current LiDAR projects that is designed to produce 1-foot-contour equivalent accuracies has 5,000-foot by 5,000-foot tiles that average 220 megabytes per tile, or 245 megabytes per square mile. Similar tiles compiled from photography consisting of mass points and breaklines are normally less than 10 megabytes.
Add the storage requirements for a typical project that includes 6-inch color orthophotos and a LiDAR elevation model, and your requirements reach 580 megabytes per square mile. Other project requirements can add considerably more storage requirements. For example, many clients today request an additional three-band orthophoto deliverable in terms of a false color or color infrared image. With these project additions, storage requirements can easily reach 1 gigabyte (1,000 megabytes) per square mile.
Luckily, the cost of portable hard-drive storage has dropped significantly over the last few years. Today, a 1-terabyte (1,000 gigabytes) hard drive can be purchased for a little less than $100. This cost equates to 10 cents per gigabyte, or 10 cents per square mile using the logic for storage requirements just discussed.
I am not trying to argue that there is no value in physical plots in the geospatial world. These obviously have considerable value in many different areas. But digital is the deliverable of choice in the mapping world. Nearly all of our clients take our digital work product and add value to the maps and geospatial information that we provide by including boundary surveys, construction details, design information, as-builts, hydrologic analyses or other critical data that are important to their end clients or, for the government folks, their constituents.
Beyond MediaThe changes in the mapping profession don’t end with the choice of media for project deliverables. In fact, the most significant changes extend well beyond portable hard drives and DVD media, yet they are very much related to the previous discussion. The most significant changes are those related to the composition of the professional staff that brings together all the high-technology mapping inputs to create these deliverables.
If we go back in time just 10 years ago, most mapping firms looked significantly different than they do today. Our largest block of employees worked in stereo compilation using the film imagery that we captured along with the associated control solution to create the planimetric and topographic maps that our clients required. A considerable area of our office building was dedicated to these employees, along with the analytical stereoplotters and softcopy computers required for this effort. The area was staffed with two shifts each day, and the amount of traditional mapping, in terms of planimetric features, mass points and breaklines, generated by this group each day was significant. For the most part, our business was centered around generating 1" = 100' scale planimetric maps with 2-foot contour intervals and 1" = 50' maps with 1-foot contours. Exceptions to these rules were rare.
Remember, this was at a time just after commercial LiDAR became viable and was just beginning to change the way we approached some of our more-complex projects. At that time, digital orthophotos had been popular with our larger client base for more than 10 years, but they weren’t yet required on every project because the computer hardware to manage such large image files wasn’t really efficient in the desktop environment. And we were still a few years away from large-format, precision digital cameras.
Today, most of the analytical stereoplotters have been abandoned and removed from the office environment. Softcopy machines are still used for compilation, but we have fewer numbers than ever before. And the office space that was freed up when the large compilation machines were removed has now been fitted with additional desktop computers used in post processing LiDAR data, generating intelligent elevation models from this data, processing high-resolution digital orthophotos and extracting valuable information from this imagery.
The personnel that previously worked in stereo compilation have gone through extensive retraining and are now performing new tasks in LiDAR and imagery processing, which are our current high-growth areas. This group is detail-oriented by nature. Moreover, their mapping backgrounds and considerable compilation experience prove very valuable in their new roles, but the tasks and software techniques are different enough to require training and mentoring on the part of other experienced technicians.
Some say the only constant is change. That principle certainly applies to the wide variety of things that we do as professionals. There are days that I would love to do one of my tasks just like I did it yesterday. But in the end, the changes in technology and processes--and the challenges that come with them--are one of the reasons that I enjoy my profession as much as I do.