Existing Ground Models - Part 3

 In Part 2 of this series I talked about the pre-process steps I take when creating ground models.  This preprocessing steps to get large datasets slimmed down to make them usable.  

In this part of this series I’m going to focus on the process I have developed, primarily around Civil3D to create validation where surfaces are created, post processed and passed through a check and review process before being released for consumption by the wider design team


The diagram above outlines my process.  On the far right of the diagram is the existing ground model file (green box), this is where we need to get to.  This existing ground model file is what will house the surface(s) that will be datashortcutted for use in other models. 

So why are we doing everything to the left?

Doesn't extra files mean more files to manage?

Why do you XML out the surfaces from the validation models?

Lets look at these questions in turn

Q. So why are we doing everything to the left?

A.  The existing ground model is the most important element of your suite of models.  The existing ground surface is referenced in to just about every other design model in your project.  The design team make critical decisions from the information contained within that ground model.  Therefore, should it not attract the same degree of attention to correctness as any other designed element?  

If we look at the MacLeamy curve below and particularly the ability to impact project and cost of design changes lines you can see that the cost of design changes increases from left to right, while the ability to impact the project decreases.  Putting this in to context of our existing ground model example, finding and error in the ground model later in the design will increase the cost impact due to impacting more disciplines and could have a negative impact on the design options that can be chosen.

                                Reference : https://www.danieldavis.com/macleamy/

We also want to consider model performance.  As I said above, the ground model is referenced in to nearly all the models, any delay on file open, datashortcut reference or validate, corridor target rebuild impacted every single user multiple times per day for the duration of the project, then this will cost time and therefore money. These are often invisible costs. 

Q. Doesn't extra files mean more files to manage?

Yes, it's a few more files to manage but you only normally have to interact with them once.  Each validation file gets created, the key steps get undertaken and the file goes through the check and review process then it doesn't get modified except if that survey data gets updated in it's entirety, for example if the ground surface is build from a draft version of a survey and the final version gets received, in which case the validation file will be revised.

Q. Why do you XML out the surfaces from the validation models?

I've been asked why I export the validation surfaces to XML rather than using a datashortcut and snapshot or datashortcut and promoting.  Well it boils down to the performance and stability of the existing ground model on the far right of the diagram at the top of this page.  When we build a surface and undertake our validation we put boundaries in place, add/remove points, add/remove triangles and simplify the surface etc. Each of these actions causes a degree of performance loss.  What we are aiming for is the best performing surface when we share/consume that surface.  Creating an XML export and importing that in to the new file ensures that surface definition lists is clean in the surface that we consume in many models.  Our validation model does all the heavy lifting and holds the records of all those actions taken


I've found this method to be the most reliable method of keeping that ground model performing well through the lifecycle of the project.  It might look like more work, but in reality there is only a few extra steps beyond your normal workflow but in the long run a bit of extra time spent structuring the data correctly at this stage will pay for itself many times over across the project as well as injecting a quality control process on the received data.  This validation process allows errors to be found then either fixed or the data rejected before the wider design team are impacted by any issues found.

In Part 4 - Master ground model creation we will look at the next step, taking those XML's from various sources, creating a composite ground model and why I always create a composite even though i might only have one source surface.

I'm interested to hear your thoughts on my process.  Feel free to post comments or get in touch directly








Comments

Post a Comment

Popular Posts

Existing ground models - Part 2

Image
Part 2 - My ground surface process As I said in Part 1 , a poor setup of the existing ground model has knock on impacts not only on the ground model opening times but also on every other model that interacts with it, whether that's displaying it or some other interaction like interfacing earthworks.  So starting off with a clean and well performing ground model is fundamental to your project. The diagram below shows the high level process I use for developing those master ground model(s).  The main problem I see at the minute the majority of users jump straight to master ground model creation, skipping the first two very important steps. So, lets talk about the "Pre-process" step.   What do I mean by "Pre-process"?  Well, fundamentally that's getting the data in to a format and size that will allow it to be imported in to the modelling software to create a surface so it can be post processed and validated. This pre-process phase predominantly applies t...
Read more

My First Public Blog

Image
For someone who purports to be a "digital" person and tech savvy, why has it taken me until 2022 to start blogging? Well, firstly I'm not a natural writer nor have I ever had the discipline to keep adding fresh content. So what has changed my mind? In mid 2019, in my late 40's, my world was destroyed when I very suddenly lost my beautiful wife of 20 years to a very aggressive cancer.  Fortunately now I'm happily remarried but one of the side effects of horrific grief for me has been some impact on my memory.  I struggle to recall things I used to be able to do easily.  So I thought I better start writing things down for my own sanity. Since started my digital journey in 2011 my role has evolved from being a day to day software produce user to a role today that has a much wider remit across technology, processes and people.  Therefore, I don't sit on the tools everyday so I have a habit of forgetting how I get products to do what I want.  Again, thought I be...
Read more

Existing ground models - Part 1

Part 1 - The Problems One of the reoccurring problems I keep seeing is issues with projects that result from poor creation processes when setting up the existing ground surface model. Some of the issues are software agnostic, these include Not undertaking a proper validation exercise on the existing ground data received, leading to inaccurate ground model Taking lidar, drone, mobile mapping or stationary laser scan data and using it "raw" automatically thinking "more points = more accurate" Not having a documented strategy for setting up the ground model so that it's easier for other models that interface with the existing ground.  We need to ensure that updates to the ground model, such as supplemental data can be incorporated without needing to go and check which surface is used with every target, profile, cross section etc. Some of the issues are more product specific Civil3D creates a companion file (.mms) when the surface point count exceeds approximately 2...
Read more