Trying to obtain survey grade accuracy from GIS data is
working from the wrong end of data. I equate the mindset as trying to get one-second
accuracy out of a three-second instrument. Yes, the surveyor could turn multiple
sets of angles and average the values to get a more confident result but it’s
still a three-second measurement providing the values, but now aren’t we
talking about precision? Repeatability? Survey
grade is capable of centimeter level accuracy (sub-tenth) and GIS grade data is
sub meter. Can you smarten up GIS grade to survey grade? If the accuracy of the information is not
there… it’s just not there.. But you can certainly dummy down survey grade
to the less accurate needs of GIS.
When it comes to GIS grade verses survey grade why not
collect the highest quality, best accuracy, most complete information? Many would say collecting survey grade data
for a GIS porject is overkill, until they want to repurpose the data for tasks
requiring survey grade data. Others would say different workflows because the
two are unique through the hardware and software used. But most would say cost of data acquisition.
You’ve heard it said “you pay for accuracy”.
Assumptions can be used as excuses to validate collecting data at a
lower cost and, in this case, that assumption is as misunderstood as the
accuracy of the GIS they defend.
It comes down to knowing what you expect from the data at
any future point and not necessarily the current task at hand. Would you use a
2 mega pixel camera expecting to get 24 mega pixel performance and clarity? This
example may be over simplifying the issue but just think about it. If you know
that in the future you may need a higher resolution image would you or could
you use a 2 mega pixel picture expecting to get a 24 mega pixel image? You can
up the resolution but there exist quality limitations. Clients sadly often search
out the inappropriate tool for the job because of the upfront cost and not
understanding the ramifications of their decisions. I ask… now what does it
cost to rescan for more accurate uses? The
cost effectiveness of H.R. 4233 “map it once use it many times” comes to
mind. Huge efficiencies occur by combining different departmental needs within
an organization and funding shared between the departments. Data should be
collected to the accuracy grade of the highest accuracy required task not the
lowest. Understanding this need not be a
legislative mandate, but it’s a step in the right direction. It starts with a client’s
understanding of their end use or possible future user needs which has to be to
be understood before the “tool” or method of collection is chosen. How many
times has a client come to you expecting to use an asset inventory data set
then later expecting or needing highway design data? Of course there is the
issue of shelf life in a dynamic world… but I digress.
Underestimating the need is not a new issue. I saw it in the
static scan world supplying architectural needs and evaluating which scan
system to use for intended end result… as-built verses detailed restoration
would equate to a 10 mm noise level scanner opposed to a 500 micron scanner. Asset
inventory verses highway design… same evaluation assessment of need should be anticipated. A
good practice is to use equipment that is capable of supplying one grade better
than what you are trying to accomplish. Any adjustment to data can introduce
error which can affect accuracy so choose wisely during the planning stages of
a project to reduce your time and frustration during the project.
We need to smarten up our clients not data regarding data
accuracy, how it affects them and how costs can be shared within an
organization not dummy down the options.
No comments:
Post a Comment