Moving towards automated mapping
We are witnessing the impact of evolving technology on multiple levels: data collection, analysis, storage, output representation. The cutting-edge collection techniques can gather much more data than ever before at the fastest rates possible. Not just the same data we once obtained, but data sets that now include new metrics such as color, intensity, even temperature. LIDAR data, for example, now includes multiple returns such as a tree canopy then the ground surface.
With richer data sets, faster computing hardware and more sophisticated software, we can conduct deeper analysis of the data. This can provide insights beyond spatial analytics. This technology evolution mandates that we must evolve as well in order to not just survive but thrive.
We are now seeing software that allows you to show an example item from a point cloud so it can “train” on it. It will then identify all other similar items. So depending on the quality of the data set, we can identify stop signs, street lights, mail boxes, parking signs, etc. Essentially, give the software an example of what to look for, and then turn it loose on its own to find all occurrences of what you are looking for, like a Google search done in 3D. A couple of decades ago, optical character recognition on images was the cutting edge of this. It is now commonplace.
We are now on the way to automated mapping products. This is the next segment that will see significant evolution. We have mastered capturing data quickly and storing it so it’s accessible from anywhere. What’s next? deep diving data mining and algorithms to extract meaningful information for further use and analysis.