It is pretty hard to pick up a technical or trade journal these days and not see an article on mobile mapping somewhere within the pages.

And for good reason--these systems are very impressive on a number of fronts and offer significant options beyond a traditional surveying or mapping approach for the right projects. But like most other tools that have been developed for our profession over the last 20 years, these systems can have both advantages and disadvantages in their application over the broad range of projects that we face during the course of a year.

There are many advantages to mobile mapping systems. First, they provide a rich dataset with a very high data density for close range acquisition. These densities vary significantly depending on a number of factors, including the driving speed during acquisition, distance from the laser to the surface reflecting the laser’s energy and the laser repetition rate. Densities measured in the hundreds, or even thousands, of points per square meter are common and therefore even the smallest of features are captured in the resulting point cloud.

Second, the data is extremely accurate. The lasers used in the high-end systems are typically accurate to a few millimeters. But this is only one component of the overall error budget. This budget grows when you add in the uncertainties for the sensor positioning gained from kinematic GPS, the sensor rotation that is measured using the mapping platform’s inertial navigation system (INS) and the measurement of the rotation of the mirror that directs the laser energy across the landscape during acquisition. Positional accuracies of 1 to 2 centimeters are possible with careful planning, quality hardware, good GPS conditions and supplemental ground control. This accuracy is appropriate for all but the most demanding of projects.

Third, mobile mapping systems can be adapted to a number of differing project data requirements. Transportation engineering projects in general, and widening or upgrades to existing facilities (where acquisition is simplified given the ease of collection from driving on an existing facility during acquisition) in particular, are often great applications. The mapping of rail facilities is another application that can benefit from this approach since the rail provides a near-perfect opportunity for driving the acquisition vehicle along the existing facility.

The potential applications do not end there. Line scanners can be added to do close-range pavement condition assessment on roadways or high accuracy cross slope analysis. There are significant applications for above-ground utilities, including electric transmission and distribution lines, utility pole location, storm sewer feature collection and asset collection for signage. Automated software solutions for pole and sign detection have become quite good over the last couple of years. And there have been significant recent applications involving the mounting of these systems on boats to acquire information around waterways including levees, piers, docks and other drainage features. Many more applications are sure to come about as the technology becomes more mature.

Finally, I do not think that any discussion of mobile mapping is complete without a discussion involving the improvement of safety and maintenance of traffic during a typical acquisition. Acquisition benefits significantly from the ability to drive the acquisition platforms at prevailing road speeds. Some boots on the ground are required with most projects to place ground control points to better fit the point cloud data to the actual ground, thus removing any system bias in the data. But this work can be scheduled for low traffic flows to minimize safety concerns.

While the advantages are significant, there are challenges that must be considered for any mobile mapping approach. First, the GPS environment for a ground-based mobile platform is subject to significant outages during acquisition. While losing satellite lock is undesirable, it is almost unavoidable during most mobile mapping applications. The reasons here are many. Tall buildings are obvious in major urban applications, but even two-story buildings can be problematic on narrow streets, blocking the view to satellites that are even 30 or 40 degrees above the horizon. Passing large trucks on roadways or driving under overhanging vegetation can have a similar effect. And overpasses typically result in a complete loss of GPS data for a few seconds.

But all is not lost with GPS interruptions, thanks to the inertial navigation system (INS) and the distance measuring indicator (DMI), which together precisely measure the movement of the sensor during these GPS outages. It is important to understand that the error budget increases with time during GPS loss, so it is always a good practice to re-acquire GPS as soon as practicable to regain a fixed integer GPS solution during the acquisition to minimize problems with the accuracy of the data.

Second, remember this is strictly a line-of-sight technology. In very general terms, if you can’t visually see the feature from the acquisition platform, then you will not pick it up in the acquired data. Neither the lasers nor the digital cameras have the ability to see through buildings, extremely dense vegetation, a large truck parked on a street or even a large roadway sign. Therefore, performing acquisition by driving along subdivision or city streets will typically result in very good detail for everything found in the front and side yards of buildings, but only those features in the backyard that are easily visible during acquisition will be present in the data. Often, multiple passes are made on roadways to minimize the “blind spots” that result from trucks or other vehicles on the roadway during acquisition.

Another challenge for some projects is the somewhat limited range of the lasers used in these systems that results in a relatively narrow acquisition swath. Mobile mapping systems use Class 1 eye-safe lasers for point cloud acquisition. The upside is that this allows acquisition to take place even during times with a lot of people in very close proximity to the sensor, which is common in urban collection, without any health risks. The only downside is the relatively limited power of these eyesafe lasers, which results in restricted acquisition swaths. Most high-end systems have ranges of 200 to 400 meters extending on each side of the mobile platform.

The final challenge arises from acquiring, archiving and completing the feature extraction with the extremely large datasets that are common with mobile mapping systems. Most systems today collect point cloud data from LiDAR as well as color digital imagery. Individual LiDAR sensors send out between 200,000 and 300,000 outgoing laser pulses per second and record multiple returns from each pulse. The measured intensity of the returns is also typically recorded. Furthermore, most systems in use today are equipped with two lasers to increase the coverage during acquisition. And it is not uncommon for these sensors to include two to four calibrated digital cameras that each have the ability to acquire multiple frames of imagery per second. The bandwidth and storage requirements for such considerable amounts of data are a technical challenge. A few seconds is all it takes to collect a gigabyte of information. Now think of running the systems for a couple of hours in a typical acquisition day.

The blending of data acquired from multiple technologies can be important for some projects. This can include mobile mapping with aerial LiDAR, or it could include tripod-based terrestrial scanners, other ground-based surveying or bathymetric information acquired from bathymetric LiDAR or acoustic technologies. This data blending can be relatively easy or significantly complicated based on each project’s requirements. A significant challenge for some projects is the combination of data with differing accuracies and data densities and how the data are treated at the edge where these data meet. Often it is desirable to blend or warp the data from the less accurate acquisition to the more accurate source at the edge where they meet. Luckily, software has come a long way in providing efficient means of bringing these data together.

For example, some projects combine the high-accuracy, high-density datasets gained from mobile mapping platforms with less-accurate, less-dense data postings acquired from an aerial platform. The end result is a dataset with the high accuracy and density where you need it the most (along the ground-based, driven acquisition route) from mobile mapping with a wider aerial-based coverage that allows you to see “behind” the ground-based obstructions to mobile acquisition. This approach also realizes the cost advantages that comes with aerial-based acquisition for the project areas outside the mobile mapping coverage.

Like so many other applications of new technology, the effects are being felt in many different ways with other sensors. We are seeing overall improvements in software, mobile data storage and automated feature extraction that have applications with our airborne systems. These improvements will continue as the technology and support systems continue to mature.