InfraWorks Sites in a Flash: Dynamic Near-Real-Time Models

KEVIN BROWN: Welcome to the session of InfraWorks Sites in a Flash, Dynamic Near-Real-Time Models. For 2024 AU. My name is Kevin Brown. I have 20 years in the industry working on all sorts of projects. Large scale subdivisions, site plan, stream restoration and solar farms across Ontario.

I've been working with and loving AutoCAD since high school, and continue with this path and design, first with AutoCAD, then land development desktop. And now finally working with Civil 3D since 2008, where I quickly became one of the go-to people for help and development, and currently teaching it at Kanazawa College. I like finding the quicker and easier ways to do things. Today that brings me to AU, to show you how InfraWorks visualization models and updates can be reduced to a few button clicks and sadly a lot of time spent rendering.

I work for WalterFedy, quickly growing multidisciplinary design firm with offices across Ontario and Alberta. WalterFedy serves a wide range of clients from industries like health to manufacturing. You may recognize some of our projects we've worked on, like Goodyear Napanee Facility or the Cooksville Go Station. We've been rated one of southwestern Ontario's top employers, and one of Canada's greenest employers. I'm looking forward for many more years working with them.

Today we'll be focusing on how to enhance efficiencies in InfraWorks by seamlessly converting large scale CAD drawings to Civil 3D models into a dynamic import models. The goal is to make this a process quick and straightforward, empowering you to produce high quality models with minimal effort. We'll achieve this by exploring InfraWorks rules and learning how to tailor our specialization to create accurate and adaptable models. More importantly, we'll look at how to apply these techniques to make our models dynamic so they can be easily updated at any point throughout the project.

Along the way, I'll share some practical tips and tricks to simplify your workflow. What we won't be covering is absolute basic InfraWorks, like understanding coverage areas or adding city furniture. However, no matter what your current experience level is, this session will equip you with the knowledge to set up your projects correctly from the start and minimize time spent in maximizing results. For the purpose of this session, I'll be using only the out-of-the-box features of InfraWorks ensuring that everyone can participate leaves with the ability to repeat this model.

Somebody might be wondering, why do we even need to use InfraWorks? I get it. This might seem like an counterintuitive, especially the very session we're advocating for its use. But trust me, it's essential. To me, InfraWorks is the unsung hero, in the development and public consultation process. Too often in our industry, whether we're working with municipalities, clients or the public, we fail to take into some patterns, fall into some same patterns. People see 3D real time models as unnecessary luxuries, expensive, and avoidable because static display boards have worked just fine in the past.

But let's be honest. How many times have we spent days on static display boards only to repeat the process every time an option changes? The public expects the displays to reflect the most current designs, which leads to even more time updating them. And then at consultations, we ask people to imagine how this bird's eye view frozen in time will translate to what they see from their window. Or how we'll feel walking down the street.

With InfraWorks we can give them exactly what they want. Better yet, with the right setup, we can provide these dynamic real time views quickly and easily.

You may be asking why choose InfraWorks over the other many visualization tools available. It's true. There are other programs out there that provide stunningly rendered visuals, complete with perfect lighting, animation and characters, and hyper realistic detail. Some companies even have entire departments dedicated to creating such visuals. But let's be honest. That level of polish is for the end stages of the project or for marketing material. And it comes with a hefty price tag.

What we need is something different. We need to show how the site evolves and discuss it with the public and with the client in real time in a way that is easy for everyone to understand. That's where InfraWorks shines. InfraWorks fills a niche that many other tools don't. It is dynamic. It can update with your design changes, surfaces, pipes and more with just the click of a button. It's built for reviewing multiple options and comparing them from any angle, even from the same angle over time.

InfraWorks is designed for large scale areas. It's not about the micro details of a single corner, but giving you the bird's eye view of an entire site from any perspective. That said, InfraWorks does have its flaws. In my opinion. One of the biggest drawbacks is there's no built-in template, meaning that setting everything up from scratch each time can be slow and frustrating.

In InfraWorks, we can easily add surfaces, pipes and structures from a handful of Civil 3D objects. But now we hit a wall. The site coverages and features, these elements cannot be directly pulled from Civil 3D data as easily. If you ever searched YouTube for tutorials and site creation, you've likely seen hundreds of impressive videos showing you how to create beautiful detailed sites. They showcase perfect parking lots, individually modeled buildings and carefully placed street signs.

Each coverage area is meticulously selected, extract from the CAD file and then individually stylized through data tab, with a specific modifications made to every piece. It looks incredible, but it probably took days to complete.

For us in consulting spending that much time on a model can be hard to justify. Even worse, when the basic design changes come through, you'd have to go back and recreate each one of those individual shapes, likely spending another day or more just to keep everything up to date. At that point, the process becomes too time consuming and too expensive, moving into the same cost territory as those high end visualizations programs I mentioned earlier. And honestly, most clients don't won't want to pay for that, and we don't have the resources to do it either.

We will streamline this process by converting our CAD line work into ready made visual models like this, complete with features, signs and trees. And this will all be done relatively quickly. To do that, the first step is that we need to move away from individualized files and move to an overall file approach.

I suspect that many of you haven't explored SDF files in depth. They aren't widely accepted by many GS programs, but their acceptance is growing. That said SDF files have their place within Autodesk and they offer some unique benefits. Like most GIS exports, you can create layer filters, ensuring that only specific information from selected layers is included in the export.

And your SVG files go a step further. Unlike many other formats, they allow you to customize what's exported from each layer. You can choose to export only polygons, just linework, depending on what you want to display. For example, you can choose to show only the center of the block from certain layers.

In addition to that you can export useful metadata from your objects. For instance, you can include the area of polygons or the scale of blocks all in a single process. But the real game changer is that you can all be done at once. Instead of creating and managing 20 separate files that need to be updated every time the design changes, you can handle everything in one file with just a few button clicks.

First we need to think things through. Now that we know how to filter and extract data, it's time to start thinking things through. We need to decide which coverage areas we want to show and what objects and buildings are essential, and what specific data we need from each of those objects. For an individual drawing, this is relatively simple. But if your goal is to create an overall master template, it becomes more abstract and can even be time consuming.

The good news is we can always update the template as needed. Personally, I like to organize things in an Excel spreadsheet like above, where I can keep track of everything in one place. For this presentation, I'll keep the list condensed.

We are first need to determine which layers will represent our coverage areas, and therefore require polygon export. For example, we look at our drawings. We have layers for edges of asphalt, concrete areas and some hatches, but we don't have a layer for the grass areas. To keep things simple and avoid manually hunting down every single grass gap, we consider the outer property line as our limit for grass areas. This will take advantage of how coverage areas can be layered in InfraWorks.

Next, we need to think about what should be represented as an artistic lines. Meaning they will be exported as linework rather than polygons. These would include things like line painting and hatches. Unfortunately, we do need to explode the hatches if we want to keep the line patterns. Line painting in particular adds texture and detail to the drawing and can be important for the overall visual impact.

Lastly, we need to consider discrete features like trees, signs, light standards, any other fixtures. These should be filtered into points so that we can display them exactly where they are.

Now that we've determined the layers to focus on, essentially review our plan to ensure everything aligns correctly. We need to convert all asphalt and concrete areas into closed polygons. It is crucial that these polygons are truly closed, not just meaning at the start and end. In some cases, this may require overlapping the line work like above, and where coverage areas are adjacent to one another. We also need to make sure it's closed, like I just said. So that's complete polygon.

Additionally, we need to make sure to explode any line hatches to preserve their patterns. Next, we need to verify all blocks are on the correct layer and have accurate names. This overall step is often the most time consuming, but it should be straightforward if your company adheres to proper layer standards and your team follow these standards consistently. Introducing hashes for certain coverage area types can help. We can also modify and reposition these polygons as we make individual changes for the site plan. As you can see here, we can take a hatch and we can extract the outer boundary really quickly.

Now that we've completed our preparation work on the site, let's make the process easier. First, we'll create and save an SDF file in a location close to our InfraWorks project, so we know where to find it. Next, we need to review the layer chart and select the layers we want to use. The layer filter carefully chooses all relevant layers, including those with polygons, outer hatches and blocks. It's better to select more layers than you we initially need. We can always filter out the necessary information later on in the process.

After setting up the SD file review, we have the future class tab. Initially, you'll see that all data is grouped into a single object, which means we won't be able to separate the data, however, by clicking a button we can break this grouping and separate the layers within the same file. This separation allows us to filter data individually. This setup enables us to refine our export by adjusting the object types and data.

For coverage areas, choose polygons for linear features, select lines. For blocks, choose points to capture the center of objects. Well, we could leave everything as a multi-selection, separating them allows us to easily filter out partial linework or miscellaneous blocks, saving us cleanup time. If you don't, you'll end up with mild linework showing up in places you don't want it to, because it's not a closed polygon.

For the third and final part, we need to select the additional attributes to extract. While we could select the entire list, I recommend focusing on the key attributes of area, block name, layer, rotation, and X scale. These are essential. If needed you can always extract additional fields based on specific block conditions. The purpose of extracting these attributes is to use them effectively in our drawing. This will allow us to control various objects in InfraWorks, managing their size, shape, and appearance all at once. For the third and final part, we need to select--

Finally under Options we need to close our polylines and add polygons. So that we don't end up with just closed polygons and them not being treated as hatches. Now that we've completed this, we completed the setting up of the SDF file, it's important that we ensure we don't need to repeat this entire process for future projects, or when adding additional information. We don't want to have to do this all over again for every project.

To streamline this, we can save an EPF file which acts as a sort of a template for GS files. This file will store all our settings, features, exports, options, allowing us to reuse configurations for any future projects. To make sure we save this EPF file in a location accessible to everyone involved and use it consistently in the project so save it next to your other templates.

Now we can see the magic in action. Notice that I'm not selecting any objects or limiting the export area. I am exporting the file. I am choosing a location to save the file next to the rest of my project. I'm going to overwrite everything. Now I'm going to load the EPF file. And you observe that it's reviewing-- you observe that it's actually selected everything automatically. All the settings are set for me.

Next, you'll see that it's exporting 1,000 objects all at once. You don't need to select anything or do anything. Essentially, I've created the equivalent of a dozen shape files in a single button click without the need to consort or confirm the data. And because I save the template EPF file, I can repeat this process as often as needed without effort.

After all this preparation, we're finally ready to use the InfraWorks. I set up the site through Model builder, imported the woodland areas from my provinces, GIS database file from our provinces and organized the various surfaces as needed. This process was quick and efficient because InfraWorks is specifically designed to easily import Civil 3D objects and large data files.

This is where SDF files and templates truly shine. Instead of importing 20 individual shape files one at a time, we'll be able to import everything from our single SDF file once. At the end of this process, we'll have about a dozen uncategorized data sources that need to be organized. The advantage of our approach is that we can categorize them in bulk rather than individually, sorting them into coverage areas, line features and city furniture. While InfraWorks processes this data file, we can see it comes up with all the other sources. We can remove stuff that we needed extract, and extract the information and places it inside our Data tab.

Now we're going to reap the benefit of a unified export and import process. Now that we've imported all the data from our SDF files, we need to take a moment to consider how to handle it. For that we can refer back to our Excel chart. As you can see, I like going back to this thing. It makes it nice and easy to figure out what you need and not have to-- and not be confused. InfraWorks has various object types. So we need to determine the appropriate type for each layer.

For this project, we're going to keep it straightforward and choose three categories. Coverage areas, line features, and city furniture to properly display the data here are the properties we need for each one. For coverage areas, we need layers and area. For line features we need just the layers. And for city furniture we need the block, name, scale and rotation. As you can see, we need the most objects, most properties for our city furniture.

First, let's select the layers and data associated with the coverage areas, such as asphalt, concrete, and stone gravel, and any other relevant categories. I've reviewed my chart and selected all the coverage areas at once so that I can apply the features simultaneously. I'll designate these items as coverage areas.

Next, I'll go to the Source tab and I'll make sure to drape the object. Without this all-- without draping, all the areas will appear below the site. And I've made this mistake a lot. Then we go to our table tab and we'll add the layer data that we were exporting earlier under the tag category.

And finally scroll down to the area tab and add area to the area field that we extracted previously. This provides additional options for separating the data. You notice that I didn't use the name category for any of this. This is because each object must have a unique name. If we use a name as a layer, all objects would have the same name, resulting in only one of each layer displaying on screen.

Now you can see all the coverage areas on the screen at once, but unstylized and clear. So we've managed to import everything successfully. Now let's look at linear feature data. This includes line painting and curves.

We'll select these features and follow a similar pattern with the coverage areas. We'll assign it as a coverage area. We will drape the object again so we don't forget. We'll assign the tag category as a layer once more. And we will assign areas area just because. This way we can keep things consistent and there's never any mess ups or forgetting.

However, this time we'll also add a buffer size to our plan so that each line will be given a thickness and make it visible on our plan. Linear features can be very widely, especially curved while often categorized curves as barrier features. For the sake of simplicity and speed in this project, I'm making them still a linear feature. So we've added a buffer size of about 0.1 meters or like about 20 centimeters. So it's a little thicker so we can see it.

Now, the final part is let's add our city furniture category, including cars, lights, signs, trees. And we'll use similar process before. Categorize everything as city furniture. Make sure we drape the information. Other words, again, it'll end up way far back down. And that looks funny.

Tag the category this time as a block name instead of layer. So we make sure we to choose block name. Instead of the-- this approach ensures that specific blocks, even if they are on the same layer across multiple objects, they will display consistently. Next, we want to transfer the physical properties.

Object rotation. Apply the rotation to z-axis. Note the block rotation is recorded in radians while InfraWorks uses degrees. Convert the data by timesing by 180 divided by pi. The object scale will match the object scale to the block scale across three dimensions. Sometimes you may need to adjust the block scale by-- these ones were 0.1, so you need to multiply these ones by 1,000.

You can see that the default inverse cubes have appeared, are starting to appear in our project as soon as this loads up. You can see them down below as we zoom in. Those are our signs. And also our signs and some other street furniture, like lighting and things like that.

Let's take a look at our site. Throughout this process, I didn't select a default style for anything. So you can see clear coverage areas in InfraWorks cubes representing the various city furniture and coverage areas. Now we have everything we need to create a site but unstylized. This is where our rules come into play.

The reason I want to use rules is to avoid manually applying a style to each data source or individually selecting different blocks to represent various object types. I also don't want to repeat this process for every model I create. Similar to the SDF file setups, rule files can be exported as a JSON file allowing you to save this work and reuse it for every future project without issues. Rules are also more flexible. Even if layer names vary slightly or there is a phase information, I can apply these consistent visualizations to those objects without needing to control every specific layer.

Just like I was able to extract an outline for the hatches, that can stay on the hatch layer and I can use one rule in order to grab both the outer line of the hatch and also the other parts of the object. So objects may update. With rules, I can selectively decide how things are applied with the same data set. For example, if I want each tree to have a different style or randomize how information displays, I can do that.

I can also qualify the data based on different pieces of the block information. So if a block says one thing or has a different title, I can display it differently within my rule set.

Now let's take a look at our rules by selecting them under display tab. You'll notice that some rules already exist, and each of these rules references the tag information. We're going to use the same format as well, just to be consistent. You saw before that we changed all our layer information to tag. The existing rules are primarily for corridors and follow standard corridor coding, ensuring that these areas appear correctly in your drawing. By modifying these rules, we can change how corridors are displayed. From these examples, we can see the rules are divided into four components.

Name. Description, where we can add additional information. We also want to make sure we use a name that's descriptive to our company or descriptive to us, so that we don't confuse it with existing default rules. The formula or qualifier defines the conditions that we need in order to be met to apply the rule. The result or the specified style or conditions are true. Note that you can provide more than one style for the outcome, and you also weight these various styles to provide some randomness to your styles. So you can randomize some of the trees and randomize the type of cars to display.

Rules can be also be turned on and off, meaning that you can load different types of even contradictory-- you can load different types, even contradictory ones, because you have the ability to select what rules will apply for this job. Maybe because you want a different style or you want to show something different, but you can keep your work and reapply it. Finally, you can also observe these rules or apply to it in specific order from top to bottom.

This means you can apply multiple rules to a single object, but only the latest rule in the sequence will take precedence. This means you don't have to make exceptions to rules. You just need to make sure that your rule is at the bottom or that it will be overwritten, so it will be overwritten by the previous rule type.

For the most part, now we're going to look at coverage area rules. For the most part rules can be easy or as hard as you make them, but they generally follow a simple pattern. What you are measuring. The object name. What you were measuring, the object name, area, tag, block name or any other criteria.

How we are measuring it? Is it greater than, less than or equal to similar to? Any of the qualifiers. And what conditions need to be met in order for that rule to be true? So let's consider what the rule should look like for our asphalt coverage area.

We want to-- we are measuring the tag that we qualify as the layer. So tag. We want it to be like. We want to use like because it will be comparing a string. And we want to say, this way, we don't want the string to be perfect.

If it's a string, we want to add the little quotations, little ampersands or whatever beside the text, and to the start and end of the text. If it's only a partial string we were looking for, we need to add more. Add in more. We'll add the percents on either side to represent wild card characters of any length. This way we can add in, like, all the layers that say asphalt, even say text or anything else.

And finally, the partial string of asphalt. So in this case, our coverage areas are going to say tag like asphalt. Tag like asphalt in order to provide the coverage areas. Anything specific. So let's put this into practice and create our coverage area rules.

First, let's add the rule. Give it identifiable name. Next, we choose the-- all right. Next, we'll add in our rule and we'll do it fairly quickly. So we want to say tag and make sure we can type it in correctly. Like, you see that the text is coming in color to show you that it accepts and knows what that these are valid points. And you can see color coding at the side there.

Put in the asphalt. And now we can hit OK finally. Next, we want to go and add in what our rule looks like. We can plus sign and we can choose anything. I'm going to choose a nice dark asphalt, but we can always improve or change the coverage area that we want. You can also get additional coverage areas off the internet, but I'm choosing to grab ones that are pretty straightforward that are in the box, just to make this easy for everyone.

Finally hit the run rules button to activate. Unfortunately, as with most regeneration, most things in InfraWorks regeneration InfraWorks takes some time. So the more rules you create at once, the less time you can spend regenerating if you're not doing a step by step. The faster your workflow will go. Now we have a perfect asphalt base for our parking lot and other features. You can see it. You can see it right up there.

Next, let's take a look at our street features. Let's make a rule for city furniture. To double check the process, we can look at the cube and confirm that the cube is, in fact, what we want and has what names it has. Because we want to make sure we get this correct and we want to do this multiple times.

To create our new rule, we'll again apply a name that is similar to it. Eventually I'll type it right. And we'll add in a formula that's pretty similar to the last one. But this time we're going to say tag. And say like. But this time, and again, we're going to put in part of the block name. In this case, LS. This will grab any light standard that it sees.

Next, we're going to choose a light standard to look like under the city features. Again, we can load in 3D models from off the internet, but we can just grab the ones out of the box here. And then we can hit Run rules. You can see that I've skipped over the regeneration, but we can see that we now have our light standards, and they're all rotated correctly as well, because we applied the object properties. We can also, if we took these properties, we can add scale and make the trees look bigger or the light standards look bigger or smaller based on our scale.

Now that we can create an all rules, we need to repeat these steps for every object in our project. You can-- as you can see in the screen, I've created rules for each of our coverage types, and add a few more options. This consensus process of updating and checking and confirming takes time. To avoid having to redo all this in the future, I'm going to export all these rules to a JSON file and save it to a secure location. This way, you won't need to recreate the rules for any future projects and I can keep adding to it. And I can use it again and again. This single file now serves as our InfraWorks template, which we can now update and improve whenever needed.

Now let's see our saved rules in action. I'm going to take the rule file I created from a previous job and apply it to this one with a single click. We're displaying the parking lot, asphalt, gravel, light standards and single assigns. However, when you look down at the model, you'll notice that something is not quite right. Once it loads up.

There we go. You'll notice that some coverage areas are missing and that grass coverage area is covering everything where the parking lot should be. This is due to the layer order in coverage areas. We can fix that by going up to layers, finding our one for the outer boundary, which we use as grass, and moving it down to the bottom. We can also rearrange any other layers we want in this process.

But we're still not completely there yet. You notice that the parking islands aren't showing up as green. This is because the islands have an asphalt outline. Most people would simply fix this manually by adjusting the islands, but if you do that, you will have to go back and repeat any and all manual edits all over again each and every time you update this project. Furthermore, changes are identified by the object name.

If you rerun the site, then the object name could end up being somewhere else on a different object. You could end up with a grass parking lot or pathway. This has happened to me a lot and left me looking a little red faced trying to figure out why random pieces throughout my project were changing on every update. So we need to stick to using rules whenever possible.

To fix this properly, I'll create a modified Rule that applies to the areas with the asphalt tag, and adds an extra condition. We'll create this new rule for islands that applies to any area with an area greater than 300m as you can see on the screen there.

So let's take a look at our standard rule asphalt rule set. We can add additional conditions and operators by selecting end or or. In this case, we are going to add end. And add a new condition so that any asphalt area that is less than 300m will apply to the new rule. Because we added, we can include the polygon area properties in our original SD file. We can use this now here by adding the new condition, end, area, less than 300.

Conversely, we could also add a grassed area to the old rule by saying or, and adding an or tag so that asphalt areas that are greater than 300 would also be applied to grass instead. You don't need to make both these rules. You just have to make sure they're correctly ordered.

Now that we have our rule set apply, we can apply to the model. Without naming the formula, we can-- Sorry. Without naming the formula we worked out and selecting the style. You know that we have added the rule to the very bottom of the list. So it overwrites any other rules, including the asphalt one. And soon you will see your islands. You got a rule in place. We're adding in grass to the process. Not grass with outline. And you can see our islands giving our project some real texture and really looking like it should look like.

So let's see what this-- sorry so let's see this template in action. To enhance our model further we're going to add site features like trees, cars and our missing building. Typically, people add these items individually, but that's very time consuming. It means that any updates would require us to repeat the process. Instead, we'll use custom blocks to represent our cars and quickly distribute them across the parking lot instead of adding them manually to the part to the model. We'll also use multiple car styles and apply them randomly, creating more realistic look for the parking lot with minimal effort.

For our trees. We'll refer to the landscaping plan and incorporate the block names into a rule. So this will allow us to represent the trees on site, adding both in depth and aesthetics to the site. You can see the site immediately starts looking better. But our cars are sideways and our trees aren't showing up quite right. We need to make some adjustments to our configuration.

Now that we applied the rules and built a solid model, you might be wondering, where's the wow factor? Where's the flash? This was supposed to be sites in a flash. Here's the exciting part. We've created templates.

We created templates for the SDF files. Establish rules we can quickly generate the site in a few steps. Run model builder to bring in the site. Import our rules from our JSON files to bring in all our coverage areas and cars and every other option. Next, we will run our SDF file in order to get all our coverage areas and objects from the CAD drawing, so we can bring it in the future. After that, oops, here we go. Bring in the STF file. Got ahead of myself there.

After that, we'll import the STF file into our drawing file so that we can get our areas. Finally, we will add coverage areas. We will convert these files into coverage areas, line types and other things so that we can categorize it all and make sure we drape it as I bring it up again. Again, just a reminder, we want to use layer names and include our areas. Click OK.

Finally, we are going to apply our rules. First fix the coverage areas. We'll apply our rules here and we'll see the site is converted instantly to our site. In total, this involves about six minutes of actual work and roughly about 10 minutes of waiting for things to process and regenerate. To update this model, all we need to do is run the STF exporter again and refresh the file information.

For different scenarios we can simply revert to the original site and repeat this process. You can see here, I'm making a quick fix. As you can see, we've completed creating our site. We have all the trees up above. We have our building, multiple cars, and everything else like that. We can now basically zoom around our site, take a look at it, do anything we want with it, and can present this to present this to the city, all by taking advantage of our landscaping plans, random cars, and our CAD file.

If we need to make changes, all we need to do is adjust our polygons, or make new polygons in order to add this stuff in. You also notice if we copy and paste in new trees and everything like that, they will also update with this process.

I'd like to thank you for attending my session. There was a lot of information there, but I'm certain that you'll be able to take your time, put together rules and coverage areas for your site, and you'll be able to have-- be able to create your site processes in about six minutes of actual work. Take care. Thank you very much.