A Workflow for the New Age of Clash Detection and Data Use

JOE SAID: OK, hello. Good morning, good afternoon, and good evening. I'm Joe Said. And today, I will be discussing a workflow created for the new age of cloud protection, focusing on cloud-based task detection and the digital control towers created from the data. Thank you for listening.

On today's agenda will go through a quick introduction about myself and the objectives of this presentation. Each section will have two parts theory, one part workflow with a heavy emphasis on the workflow to give listeners the ability to duplicate the workflow on their own. This includes the cloud-based clash detection, creating digital control towers, a.k.a. dashboards, and implementing the work on a project and company level. I'll conclude with a quick summary.

Before I get started, I'd like to give you guys a quick introduction on who I am and why I'm giving this talk. Quick note that BIM and data stars, just show how much I use BIM and data in that role, not how effective it is in that university or [INAUDIBLE]

So I started with a bachelor's and master's in construction and civil engineering. Coincidentally, I almost got a minor in data analytics thinking I'd leave the AC industry. But instead I jumped right into the field. And although I didn't get to use BIM that much or data in the field, I think the experience I got there is invaluable, and any person in the AC industry should spend some time.

After a bit in the field, I ended up doing a specialized master's in BIM management, which really elevated my BIM capabilities outside of the software. During that time, I worked for Tecnimont, and a huge shout out to Enrico de Maria for opening my eyes to the ability to integrate data visualization and BIM data. Most recently, I worked for Buro Happold, where I was able to work in BIM specific role and combine all these experiences into this workflow.

So if you read the three objectives of this presentation, they aligned with this slide, hopefully, which is to create a clash detection workflow to set up digital control towers and to implement this seamlessly. There are six key points I hope everyone gets before I conclude, the clash section, we must learn how to get useful data for ourselves to overcome overload of clashes and how to set up-- yourself up for locating elements in the model so that the engineers can locate the elements in the model as well.

For the digital control towers, we have to make sure our data transfer is clear and clean. We have to focus on our data pre-processing and cleaning and create an automation schedule in sequence. By combining these two objectives in six key points, we get the entire picture of the implementation. We'll also talk a little bit about the non-technical challenges I faced while implementing.

Before jumping into the meat of the presentation, I'd like to show you the presentation format, which makes it a lot easier when you're following along. So our six key points will be highlighted during the workflow slides. They are highlighted whether it's something that's a problem I face or the solution for it. Next to that you may see an A, M, or both.

As we all know, we're not always the admins of the Autodesk Construction Cloud Project Hub. So I will show the workflow for both admins and members. Finally, you can see the agenda on the right.

First off, clashes. I won't spend too much time here since I assume most people watching this will know about what a common data environment is. But in case you don't know, a CDE is a centralized digital hub for all project-related information. The ideal goal for of all CDEs is to allow all members of a project to go and visit one location and get any information they want for the project. Autodesk Construction Cloud is an example of CDE, and is the most widespread used one in the industry.

Before the introduction of the issues interface, cloud protection was primarily done and recorded outside of the CDE. And although there were some plugins that allowed for this data to be centralized, they weren't as seamlessly integrated into CDEs. Now, with the issues interface, keeping the clashes within the CDEs solidify the centralization of project information that allows for further ease of access.

So clashes have been around since the beginning of construction. We've had clashes from way back in the day. But now, we have a new old age era and a new age era. Before it was more isolated. Well now, we're trying to get more integrated.

We used to use older models that were slower and needed to be-- the Bim team would download the models, do the clash detection, and send it to people. Now it's faster because it's now online and available at any time. Again, like I said, the current model allows you to do clashes right on the models that are live. And since you don't need to download the models and do the clash detection on your own laptop, you can do it on a cloud-based standard.

Before, it was hard to share because you'd have to wait for the BIM team to finish last report before sharing it. And now, as soon as you create the clash, it's in the hands of the engineers, which increases the visibility across all project stakeholders.

Now to get to the fun stuff. First, the CDE setup for me. Just as a reminder, I'm going through the workflow of online clash protection. We will discuss how we will make sure to get useful data and set up the CDE to export the data, how to tackle an overload of clashes, and how to help engineers in locating them on.

I think I can skip over some of the ways to set up the models, set up the hub on ACC, which is an admin job so I can get more time on the clashes and data visualization. But if you have any questions setting up the project, please reach out to me. We can chat later or you can check the setup from Autodesk Docs.

Setting up locations is one of the first crucial steps I like to do. This is for the admins only. Go to Settings, Locations, and you can set up from an Excel or from the Revit. You can also download the templates os that you can know how to set it up. I recommend setting up your locations based on what you want to see. So for example, it can be different buildings, it can be floors, or it can be rooms. The granularity is up to you.

Another important thing I don't think to cover in depth is creating the coordination space for admins. Only the admin can create it and set it up for the entire team. This type of work makes sure that it's flash on and set whatever folder you want to set it to.

So after setting up the model, you create the locations and coordination space. You get something like this. This is relatively complete and clash-free model that the Autodesk has given us. This is best known in template. But still, there's multiple clashes-- thousands of clashes that you're not sure about. So we're going to go over how to set up some of the views and the filters to avoid and organize these clashes.

First off, you go and create some views. In the model, you select the models you want to set. I wouldn't recommend putting them all into one setting so that you can take it bit by bit. So for here, I did the plumbing and structural. You go inside, model browser, you set the disciplines you want to work on. And you can set more advanced filters. For example, I like to set my work set so I can just do, for example, drainage on its own, ventilation on its own, or domestic on its own.

As you can see, I just saved the view. And then, once you close out of that, you'll see that you have multiple views. I have views for mechanical rooms, structure versus different things, the mechanical models together, MEP. And when you go back in, you'll find your clash-- your filter still in place. This helps a lot when prioritizing what you want to coordinate first.

After creating your views, your clashes might be a little bit less large. As you can see, the structure versus drainage is now only 249, which is how far to start with, and then go back and do other breaks.

Now that you have your setup ready, you need some issue templates. I personally don't like to create issue templates, but this is a template that saves you some clicks. You can set it up from the issues, create new issue template, title, you can fill out any of the fields. Generally, I like to leave the description blank specifically so we can find the element in Revit.

So before getting into it, I'd like to talk a little bit about the field and how I use them. Every issue has a number and a issue thumbnail. The thumbnail is just a screenshot for when you create that issue. Then, you have your title and status. Title can carry the location or element ID.

If, for example, you are a member, you put-- I like to put the level before in the title just so I can extract it out when I'm using the data. And then, the status is the step that this issue is on. Usually, you set it up as open and then the engineers and yourself move it from pending to in review to closed. There is more statuses available that you can set up.

Next, you got the type, which includes the category and type of clashes. Again, this usually is auto-populated when you create your clash and it's customizable, which allows leadership to understand if it's design problems or coordination problems or anything else.

And then, like I said, the description auto-populates. As you can see, the description here populated the elements we have. Then you have the assigned to the watchers. And the urgency assigned to is what's assigned to, watchers is who can see it, people that are involved, and the urgency is a numerical value to help the team rank the clashes.

The rest of the fields include the location, location details so if location isn't set up and you're a member, you can put, your location, information location, details. The due date, which is expected date of resolution, the start date, which is when you should start, and the placement root cause references and comments. Placement is what model you clicked on when creating the clash. Root cause is pretty similar to type. I try not to use it. And references is pictures or PDFs. You want to give more information, you want to have a conversation.

And finally, it's the comments. I suggest to always use the comments because it's easy to communicate with your engineers. You can use the comments in Revit and in ACC,-- and it gives you a chance to communicate on the without waiting for a response.

Now, you're on the last step of creating the clash issues. You go to your clashes, you select your models. If you group by something else, I don't recommend it. When you group by object, you have that element ID. You zoom in, you find your element, you find your clashes, or you can go one by one. Here we have a clash with a T drain and a beam.

Make sure to click on the model you want. So not the green because I'm working with plumbing right now, and it creates the issue. And at this point, you're ready to fill in all the fields we just talked about. And as you can see, the element number is there because we grouped by object. This is very important when finding the elements in Revit.

Now, I'm going to get to Revit. In Revit, once you download the issue, the add in, which is now during the recording of this presentation is only 2022 to 2024, you'll have the issue ribbon, as you can see. You click on the issues and you map to your associated model. You might get this error, which is a reminder that you need Autodesk Desktop Connector App and to select the project in order for this to work.

Make sure to add the project in the desktop connector app to select the models in the shared folder. So for example, here, sometimes I do my clashes in shared, sometimes I do live. I like to have both of them associated to my interface so I have all the clashes across.

And then, at this point, you have all your issues on the right. Once you get that, you can navigate to your 3D plan and section views. And since we already set it up with our element ID, you can select the element ID, go to the Manage, Select by ID, and isolate it. The next step is to resolve the clash and to close out that issue.

That's it. Once you do it, once it gets easy and seamless after that, I always encourage the engineers to use the comments section so there's a two-way path of communication. Now, hopefully we got this so that we're set up with the clash detection. We don't need any other software. The communication is seamless in the same modeling software that you're using to create the models.

Remember that we need necessary data, useful data, we need to standardize the data input, we need to think of our outcome in early. For an overload of clashes, make sure to compartmentalize your by priority. You can create your filtered views. You can prioritize clashes. And then, finally, we help the ease of locating elements in the model by communicating with our engineers, making sure to tell them where they need to look for our elements and to utilize the autocomplete feature.

Now, for the digital control towers. This is just a fancy name for dashboards. Before we get into it, I have done some patch research focused on big data management where I came up with eight steps to help in this process. First step is exploration or understanding the data available, mapping the source of the data to the expected outcome, extracting the specific data to be used, formatting the data to be suitable and clear, visualizing to tell the story of the data, automating the extraction, updating and uploading of the dashboards, and then testing it with your end user.

And then, you reiterate. It's very important to know that this is an iterative process. You're not going to get it right on the first time. You got to try it with your users and see what works for them and what works with you.

Some of the things I want to emphasize is formatting. So for this, I'm going to use Power BI. Formatting is very important because in Power BI you have three different ways to format your data. First off, in the transform data in the M language, in the ETL, they call it, Extracting data, Transforming, and Loading data.

The M language is used for more complex processes like merging data and combining data. That's it. And then, in the model view, you link the data. So this is kind of similar to the mapping. First you map on a piece of paper or in your mind. And then, when you have multiple data sources, you want to map those parameters to each other. So if you have username on one side and username in a different table, you want to make sure that these are linked to each other.

And finally, in the table view, you use DAX language, which is also similar to M language, but not the same. And it's more for data modeling and analysis and real time analysis.

Now, another thing I'd like to emphasize is the automation. Usually, to automate the dashboard, you need to two steps. You need to automate the exporting of the data and you need to automate refreshing the data. The level of automation is, again, from 0, 1, or 2. As an admin, the goal is to always have a 0 where it automatically refreshes every week on its own. As a member, unfortunately, you have to be semi-automated because the extracting isn't as-- isn't allowed to be automated.

As you can see what I just said, the admin has the data connector from ACC. And due to Power BI being a Microsoft product, we can automatically publish the dashboards to Team and PowerPoints, SharePoint, anything from Microsoft.

Now, the member doesn't have that same option of data connector from ACC. Instead, they'd have to export manually to Excel and then refresh the Power BI the same way as the admin. So here, I'll talk about the data transfer.

Again, our three key points. We want to learn how to transfer data to and from Power BI. We want to be able to understand the process of data processing and to implement an automation strategy. Extraction, you must be sure to extract if you're an admin with all. I apologize. If you're an admin, you want to extract directly from data connected.

First, you run your extraction, and you got to make sure-- if we're doing the issues, you don't want all the data. You can get the admin, the issues, the location, and the relationships, or whatever is required in the template you're going to be using.

The next step is to export using the Power BI templates. This is something that ACC has created with multiple sources. This allows you to, basically, do nothing. You just download their data template and you use it in Power BI right away. For example, here's the issues dashboard. Download it, connect temper, and you open it.

Once you open it and you've already run an extraction, you'll see you connect your account, your output, and your project name. Once you do that, you sign in and you connect. You'll see that it'll load all the data that you selected. So for me, it was the issues, admin, location, and relationships, and it's pretty much already there. Everything's there. You don't need to do much.

Although it's a little bit convoluted, it's useful. If this is what you need in your team, it's an easy step and you're done. But again, this can only be done as an admin, and the benefit of this is it also works on a project-- like an enterprise level. Or if you want to be more customizable, you can still connect to that extracted data by connecting to the Autodesk Construction Cloud from RBI. You connect by getting your data and selecting your project and selecting the data you want to import.

Again, it'll load it all in, but this time there won't be any template. This is where you create your own visualizations, you fix your, data, and you combine the mapping together.

Now, the member doesn't get all this fancy stuff. They have to go through the docs and issue template. This is where you export Excel as a report of all the issues you already have. I usually like to date it and say the project before it. And this is the Excel that you're going to use that you're going to run maybe weekly. So every week, you'd have to come back, export this data.

This is what it looks like, basically, in Excel of every field that you have for every clash that you have. And then, you import it pretty similar, import workbook. You select it, if you want to select just one project to do for one project. You load that in, and then you'll have just one table. Not like the admin way, but still available-- everything that's on that clash report.

As a member or something across all your projects, you can import a folder. This folder is the one you saw earlier that has project one, project two, project three. When you import a folder, you got to combine and transform the data so it makes it one big table from the folder. And it's always based on a sample file. The first file, usually, works since we're exporting the same file across different projects. And then, this is where the Power Query and the M language comes up. You'll see that you'll get everything from the source name and then you'll have--

This will take a little bit more manipulation. You want to extract your project name. You want to make sure that your IDs don't overlap because they're different projects. It takes a little bit more effort, but this allows you to look as across your entire project-- an entire enterprise of projects as a member.

And then, it ends up being pretty useful. For example, this is the one I use for our individual projects. It gives you the location, the discipline, the issues, and which are open and which are closed. Pretty simple. Exactly what you need. Anybody from your team will know what they're looking at when they see it.

And then, if you have an enterprise, you can, again, set it up by project, by discipline, you can look at your total number of issues across the entire team. This is, again, from the folder and a little bit of manipulation. I'd want to get into the manipulation and the extracting of the data, but that's more of a Power BI thing more than a connection between power BI and ACC.

Now, I'll try to talk a little bit about some of the software solutions for Power BI. Sometimes, you get that template and you're ready to go and you don't want to do that much work and you see this right away. So this is where you need to go into your transform data. And when you open it, this just takes a lot of time to get used to. There's going to be some times where it's going to be a simple fix like, oh, the name of the project isn't right. And sometimes, it might be a complex problem like you're extracting isn't correct, or you need to recreate your extracting procedures with a different sign in.

On the left, you can see every table that was extracted. And those little exclamation points are ones that need to be fixed. You go one by one, you try to fix it. And what I usually do, if I have a question that I can't solve, is get on the Power BI forums, one of the best resources out there. A lot of people are they're willing to answer questions. And I usually get-- if I have a question, they'll answer me right away. And eventually, you solve all your problems, you get everything done, and then it works.

Another problem you might find is relationships might not be set. You have two different elements from two different tables that need to be connected correctly. Or that's when you go to the linking model. For a small project, like the one I made, there's not that much. You have the enterprise folder and I have a supplementary Excel sheet that has the user discipline.

Sometimes, it gets a little bit more convoluted. If there's a lot of data that you want to get involved in and there's a lot of tables, you need to connect everything to each other, especially the one-- the data that comes from ACC. Because, for example, the admin has all the usernames, but they sure don't have usernames, they link to the admin ID. So you got to connect that to the admin. That's why you have to export the admin whenever you work. Little by little, you get used to it and you'll know how to fix some of the problems you're facing.

And then, the final thing is using the real-time analysis by using DAX. This is where you can create new calculations. For example, here I have number of open clashes, which counts as we're going as is getting updated. So it doesn't need to refresh the data from the M language from Power Query. This is in the data table view, number of clashes.

And you also, like I said, I have a supplementary table that I created within Power BI to create a calendar just so I can have a linear process of which days late, when things are late, things like that.

Now, for the automation step. Again, going back, you remember the admin gets to automate the data connector and the refresh-- automatic refresh. And then, remember, it has to export manually and then automatically refresh from Power BI.

So this is repeating this, but you can schedule the data connector. You can it every week at a specific time. I prefer times at night when I know nobody's working, nobody's getting into it. And then, for the member, unfortunately, you just have to run it every week.

Now, for entire BI to automate this step, both the member and the admin need to publish. And usually, you'll have My Workspace or you can create different workspaces for different projects, and you select and you'll find it online. So this is Power BI online. This is what I mean by automating and being able to be visible to everyone.

You have your Power BI online, navigate to your workspace, and then you have both your dashboards. So this is the admin dashboard that I made and the member dashboard that I made.

Now, how do I refresh? You can go in and refresh it every time you want to. Just click the Refresh on the semantic model. Semantic model just means the data, and that data means that it's reading from the source. So every time you refresh it, it'll go back to the source that it's reading from. For the member, it'll be from the folder. And for the admin, it'll be from ACC right away.

But if you're refreshing in the admin, you must make sure that you're connected. You sign in and it has to be able to sign in to ACC on offline hours. So that's one of the things that you have to make sure you set up before you set it to schedule to refresh.

If, for example, you have data that's not refreshable, like RBI can't find it, it'll show you. It'll be grayed out. You're not allowed to refresh it. You're not allowed to schedule it. But once you make sure that all your data is available for Power BI to reach it, you can schedule it. I would schedule it a little bit after the data export and maybe on the weekends.

Finally, I'd like to talk about the objectives and key points again. So as we said, the data transfer, we need to export the data correctly, understand the source of the data, and import the data efficiently. Now, the data processing, you edit and refine data for clarity. This I didn't get too deep into because there's a plethora of problems, but you can-- if just takes one step and another and then you'll solve it. You can map to multiple sources.

And then, it's very important is to pre-plan for visualization. Don't take data you don't need. And then, for automation, you have your two steps, automate the export and automate the publish and refresh.

Finally, we're talking about implementation, which is sharing and utilizing the information with your entire team. So the first thing is visibility. You want enhanced collaboration. And this needs real-time data sharing clashes. You need information to be relevant. This helps unify the view between the team members and make sure that they're all aligned. You want to have proactive problem-solving. So early clash detection resolution on a project gives your team the ability to see the clashes ahead of time, see it early, and it gives your team-- management team, the trend analysis to see, hey, our plumbing team's a little bit behind. Maybe we should put more hours towards our plumbing clashing.

And then, the best thing about it is accountability and transparency. It gives you the chance to track the progress of disciplines across projects, offices, and Team, and it allows proper hours to be allocated to the project.

Buy-in. So this was actually way harder than I thought. I thought, I'm going to create this amazing dashboard, everyone's going to love it, and that's it. It's going to be used right away. But it didn't work that way. I suggest always going with the first trial. Start with one engineer that's going to optimize with you. That engineer or that worker is always going to be able to have a conversation with, hey, is this working for you? Is this not working for you?

And you don't need to do it across all disciplines, across all projects right away. Just do one dashboard with one perfect data set and show it to the project lead. Get their feedback. Again, this entire process is iterative. I didn't come up with this in one night and, bam, I got it. I tried one thing, I tried another thing, I tried many things.

And then, the next step would be to do it on a project, to develop the first dashboard, implement better visibility and communication. And then, take the process and repeat for other data sources. If you have it for issues, why not have it for RFIs? Why not have it for models, model location, shared model location?

And then, if you get buy-in from that, then there you go. Now you have company-wide buy-in. You can look at admin rights, enterprise rights on ACC, create an understanding of project performance across company. Once you have that, now you're looking at your entire company at once and you're seeing, oh, this project needs more hours, not just the discipline, but this project.

And it's useful for owners and a project. So this is usually architects or actual owners. When you're in a consultant or someone who's just hired on who doesn't have admin rights, this gets a lot harder. But you could still do it, as I said, you can export your Excel, you can export even data from Revit, you can export your schedules and do this.

Now, the workflow for visibility is pretty simple. The best thing about it, it's a Microsoft product, so you can share it to PowerPoint, you can share it to anything that's Microsoft. So again, you would navigate to your workspace, you click on the Report this time, and you get to see the report online. And you can export PowerPoint PDF. PDF can become stagnant. But on PowerPoint, you can export it on PowerPoint and you'll embed live data and you just copy and paste it. There's a little tag inside PowerPoint that says, insert Power BI.

And you'll see here, unfortunately, it doesn't work when you present it. So I have to take a screenshot of my PowerPoint to show. But I'm sure that's something that'll be fixed in the future. You can also share this to Team. As you can see, back into the workspace, you click on Chat and Team. You can copy and paste it, share it to somebody, and they'll be able to open it.

I've used this on project Team where I share it on the entire team. So as you can see, you open it, you'll have it right there in your team's chat. You can also include it as a tab in your team's chat. Click on the plus, click on Power BI, and there it is. And once you refresh and automate it, it refreshes and automates into your Team and your PowerPoint. So that live link between Microsoft Suite is very useful.

Finally, I'd like to conclude with just going back through the objectives and key points. So clash detection, make sure you get useful data, only the necessary data, standardized data, input early, and think of the outcome early. The overload of clashes compartmentalized by priority create filtered views, prioritize critical clashes, and then, for locating elements in the model, communicate with engineers, utilize autocomplete features.

For the digital control towers for data transfer, make sure you export correct data, the standard data source format, import data efficient. And for the data processing, edit and refine the data for clarity, map multiple sources accurately, pre-plan for visualization. Finally, for, automations, your two steps, automate the export, automate the refresh. And combining that together, you get implementation.

Hopefully, this can be helpful and this is something that you can use in your everyday workflow at your team. Thank you.