The New Work

In my last post, I setup my ideal organization into Production, Management, and Consulting divisions. Each segment can exist as a business in its own right and I will detail my ideal methods for dealing with each in their separate posts. The first one to be addressed is where everything starts: Production.

Ever since its inception, surveyors have dealt with spatial data as sets of discrete points that represent a given region. This is still the best option in boundary surveying, but for anything else modern technologies offer a better solution. Specifically instead of building a dataset point by point, it is a better solution to capture the region via remote sensing like close range photogrammetry and terrestrial and aerial LiDAR. The two biggest hurdles to implementing these methods are their increased complexity and cost.

There is no way to get around it; it’s time to go back to school. Not necessarily collage, but certainly the official training sessions put on by all major equipment manufacturers. Actually capturing the data is much simpler than using a total station, as long as the proper references exist. Where the difficulty lies is getting a useful product out of a point cloud. There are two options, depending on the level of computer science expertise available:  plug-ins and programs. If there is a decently experienced computer tech available, most CAD programs (including inexpensive ones like Google’s Sketchup) have pre-built additions that can be downloaded and tweaked to extract the nessicary data. The actual extraction is still fairly simple, but where the computer coding knowledge comes into play is when the lines need to be classified based on some criteria. Clear communication, patience, and solid understanding will result in a custom solution that creating the needed data. Not having though things will only create a mess, which is why there are dedicated programs for completing this task. These are going to be more expensive, but they are going to work every time with minimal training. Besides increased cost, these packages are very specific as to what applications they are good at. Mines, roads, topo maps all have their own programs that can be bought and added to a base CAD system. Carlson and Trimble both have excellent setups for this option.

The other major hurdle in using point clouds and models for survey data is the high cost of the equipment. A robotic, imaging total station is about $20K. Add an integrated GNSS receiver and it’s about $35-50 thousand depending on the model and precisions. The cheapest scanner is at least $50 thousand, and that’s without the special software mentioned above.  One option that is becoming more appealing is close range photogrammetry. Systems like Photo Modeler cost one or two thousand dollars, are very easy to use and effective at creating the same kind of point clouds that scanners do. The field setups and data acquisition are more susceptible to the environment and user error, but it is possible to create similar precisions to scanners by getting a good camera and taking more time. This method isn’t going to be truly viable until commercial UAV’s come available to capture data from better vantage points and over larger distances. Giving that this will happen in four years now is a good time to start practicing.

My preferred method for acquiring spatial data is to actually buy it from someone else. Yes, LiDAR systems are expensive but once obtained are capable of producing vast amounts of extremely precise data over large areas quickly. Instrument resellers have found they can produce datasets to sell more readily than selling the actual equipment. While unorthodox, the automated nature of LiDAR means the information is produced to a very exacting and repeatable standard as to be just as trustworthily as sending out a traditional crew. The cost of using LiDAR data drops from a million dollar setup to a few thousand for processing software and a computer to run it. Giving the unique nature of companies that can afford LiDAR data, getting them to sell their datasets may be difficult at first but not impossible. Have the local surveyor’s board hold a meeting to introduce the local professionals to LiDAR and point clouds by asking a company with the technology to hold a presentation on what they do. Networking has always been important for us, and now may grow to be on par with reputation and ability. 
Posted by Andrew Gaiennie on Feb 20, 2012 9:44 PM EST