Internal accuracy vs external accuracy
Posted: Tuesday 7 May, 2013
When we discuss data accuracy we are typically talking about how closely values in a database match what exists in the real world. In the past this meant querying a database and seeing whether the value that was returned matched what was actually in the real world. This was a simple and obvious task as the stored data was always viewed as text.
As databases have progressed over time, so have our techniques of visualizing that data. For example, spatial data instead of being displayed as tables or spreadsheets, was displayed as maps in either raster or vector formats and while it was only a graphical representation, it was still extremely accurate. Then things changed and we entered the world of data mash-ups.
To increase user comprehension, GIS data was used as an overlay on top of georeferenced aerial photography (like exporting to Google Maps). While this dramatically increased the users ability to interpret the data, it now meant that data not only needed to match the real world, but it also needed to match the other datasets. Users viewing the object data and the underlying image noticed that they were positioned differently. The user then had to determine which, if any, was the more accurate dataset.
Using augmented reality to visualize real-world objects means this trend is set to continue. When we display GIS data in augmented reality we rely on the accuracy of the hardware (compass, GPS, accelerometer and camera) in addition to the accuracy of the underlying data. Once again the user finds themselves in the position of having to determine if the variations they are witnessing are an issue with internal accuracy (the database) or external accuracy (the hardware).
In my experience, most users simply ignore these inaccuracies. Perhaps they figure it's just a glitch or that in the end the computer will figure it all out. Either way, it is critical that as augmented reality advances and becomes a more widely used technology, we must understand that the users default mindset is to trust what they see on the screen. The responsibility, therefore falls on developers and those who capture data to ensure that what the user sees is always the most accurate representation possible.
Posted by Jason Churcher
Tags: GPS, future, hardware, mobile, phones, tablets
Please check back often for updates from our management, development and sales teams as we share what's been happening with Augview both internally and with clients.