This is my third post about GIS data quality – in the first post I talked about the ever-increasing importance of the quality of your GIS data, which is necessary to optimize today’s sophisticated, complex utility applications.  I also discussed, in general, the concepts of Quality Control and Quality Assurance.  In my second post in this series, I described the components of a successful Quality Assurance program.  Today I’d like to talk about Quality Control, which I previously described as “doing things in-process to produce quality results; doing whatever can be done to force quality into a process.”  By its nature, effective quality control is proactive.  As someone responsible for the quality of the data in your GIS, taking the initiative to introduce process improvements and to increase your team’s quality focus is key.

At SSP Wind Lake, we have been GIS data-centric for going on 15 years now and during that time we have seen a lot of things, both internally and from our customers, that have worked well in improving data integrity and shining the spotlight on quality.  I’d like to share a summary of some of those things with you here today.

Measure, Measure, Measure

This is a reiteration of what I discussed last time, namely implement a consistent QA program.  For all your processes that manipulate the data, particularly the manual ones, score the quality results whenever possible.  Make sure that the feedback from your QA inspection gets back to everyone involved in the process.  Keep score by error type, so you know where to focus most and provide specific error examples to the individuals on your team.

Involve the Field (part 1)

Staff in the field experience firsthand the quality, or lack thereof, of the GIS data.  Encourage field personnel to provide redlines of mapping issues that they discover.  Build a corrections process that is timely and involves closing the loop with the individuals that initially identified the errors to make sure that the corrections were done properly.  I believe this is very motivating in getting those field staff to stay focused on quality and continue to identify errors as they discover them.

Several years ago, we assumed GIS mapping responsibility for a customer in a given area of their service territory.  As part of the overall mapping/posting effort, we were responsible for completing records corrections.  When we first started, there were a stack of corrections that were old, years old in some cases, but very few recent corrections from the field.  We dedicated resources to completing the old corrections and made sure that the individuals in the field who first identified the issues were notified of their correction whenever possible.  We also encouraged these individuals to “keep it coming”, even though they hadn’t really seen results over the last couple of years.  It began to work.  We started seeing many more corrections come in from the field.  We also started getting lots of “thank-you’s”.  I believe this feedback loop kept our staff, as well as the customer’s field personnel, focused on GIS data quality.

Involve the Field (part 2)

The source for most GIS data updates are work orders generated by engineers and field designers.  Experienced mapping technicians become very astute at identifying missing or incorrect data within those work orders.  Whenever possible, unless there is an obvious answer, errors within work orders should be brought to the attention of the field designers that generated the errors.  After all, this is real QA feedback for the designers.  If this feedback is provided constructively, I believe it leads to an increased focus on quality by both the field designers and the mappers.

On one of our mapping projects, our GIS mappers interact directly with the customer’s designers.  This was the doing of the customer’s GIS Manager.  She had her mapping staff working this way and she wanted us to operate likewise.  I must admit, there were times when this was a little tenuous at first, but over time this has evolved into a very efficient and effective process.  Feedback is provided, questions get answered quickly and the mapping gets done as it should.  Not to mention, excellent working relationships have been established along the way as well.  We had to work with some of our staff on their technique for asking questions and providing feedback, but everyone has grown because of this process.

Involve the Field (part 3)

GIS mappers will make the best judgment calls if they have a solid understanding of what they are actually mapping and how the network works.  A great way to build this understanding is to conduct periodic field training sessions for the mappers, at various job sites where construction is occurring.  In this way, the mappers can see the different equipment first hand, see how equipment interconnects and get a sense for your organization’s construction practices.

My next-door neighbor happens to be a Supervising Engineer for our local utility and he has been kind enough to conduct field training for our staff on a couple of occasions.  The GIS Manager at one of our customers has been employing this training technique as well, with field visits occurring approximately once per year.  SSP’s staff assigned to this customer have been invited along.  The results of these sessions have been outstanding.  The mappers have learned a lot, they are more in touch with the reality of what’s happening in the field and just as significantly, or perhaps even more so, they universally feel like their job is more important.  Win, win, win!

Specifications and Process Documentation

Say what you do, do what you say.  It’s important to maintain up-to-date specifications for your work and documentation of your overall mapping process.

Your process documentation doesn’t need to be long, in fact, concise is almost always better.  If the documentation is too long, people tend not to refer to it when they have a question.  Checklists can be very effective, as long as your staff uses them.  Specific findings from your QA process should make it back into your documentation, especially if they can be incorporated into a checklist.  If QA feedback for an individual demonstrates that they are not consistently using their checklist, this point should be emphasized to that person.

Feature specifications typically cannot be done concisely, but if they are well organized they will still be valuable.  We find a relatively simple spreadsheet format to be useful, with a tab for each feature type.  Each feature (tab) has an attribute list with relevant notes/information for the attributes to be captured during mapping.  Additionally, graphic depictions of the feature in various configurations are also included as placement examples.

In some instances, within your specifications, you will likely run into situations where there is more than one correct answer.  In these cases, it may seem most prudent to stay flexible, let individuals choose their own preferred method or approach.  Our experience, however, is that this flexible approach may cause problems down the road, as two or more staff with different approaches may over-debated the correct way to handle a situation.  By defining the preferred approach, even in these less important situations, you can eliminate questions and inconsistency.

To be most effective, process documents and feature specifications need to be proactively maintained.  We find it effective to assign an “owner” to each of these documents.  The owners have the responsibility to update the documents as changes occur in the model or mapping practices.  If you can solicit a volunteer for this task, great, otherwise you can consider making the assignment for a period of time and having this assignment rotate through some or all of your team.

In addition to specific practices such as those listed above, continuous improvement is fostered by a day-in, day-out quality-first attitude on the part of you and your staff.  Ideally, QC encourages a way of thinking on the part of those involved in a process: how can we make this better or more efficient?  Can the process be improved so that I get what I need sooner?  Can documentation be produced to make our results more consistent?  Quality Control needs to be an ongoing focus within any process for that process to produce optimal results.

“Quality is never an accident; it is always the result of high intention, sincere effort, intelligent direction, and skillful execution; it represents the wise choice of many alternatives.” – William A. Foster