How to connect Tableau and Exceler8 to visualize building information

One of my favorite tools for visualizing building information is Tableau. All you have to do is give it a data source and you are able to very quickly visualize it. I’ve made other posts here about it…

In this case, I’m using our Excel link to Revit, Exceler8 (aptly named by our own AYBABTM) as my main source of data since it is very quick and I can get more information that I can directly from Revit schedules. Also, Tableau can connect directly to an .xls file so it is convenient on that side too.

Exceler8

I’m going to export the category ‘Rooms’ since that’s what I’m looking to visualize. The objective here is to have model data directly drive some graphics that I need to put into a project document that is not model or drawing based, probably in Word or InDesign.
Exceler8 Interface

Make sure to select the ‘As Text’ option on the Parameter Value Configuration, otherwise we’ll need to do a bunch of clean up in excel which we want to avoid. Also, make sure to grab both Type and Instance parameters.

Tableau - 01

Over in Tableau, you want to Connect to Data and select the file you just saved out. The default option will be to select a single table (workbook) within that file.

Tableau - 02

This would work fine if the first row of our excel file had field names, but Exceler8’s field names are on the 8th row. We’ll need to use the Custom SQL option to call a particular named range, or in this case to explicitly call out which cells we want to bring over. 

Tableau - 04

You’ll notice that Tableau has already written code that will bring over whatever table we had selected before we switched the custom SQL option. This is a really nice shortcut to learn how it’s trying to understand what’s in the source file and how it’s trying to bring it over. In this case, it’s reading the column headers as field names which we know is incorrect. 

Tableau - 05

 

What we’ll have to do is select all and delete the query that Tableau wrote for us, and paste this in instead:

SELECT *
FROM [Rooms$A8:Z500]

What this will do is to select all fields (*) from Rooms starting at A8 through Z500. I chose 500 as a number that was obviously larger than the number of Room elements in my model. We then check the ‘Preview Results’ button to test our query and make sure it is working. 

Tableau - 06

 

The result should mimic what we have in Excel. So, without any reformatting at all, we were able to drive down into a specific range of cells in Excel for our use. By selecting the ‘Connect Live’ option, any updates to the data that happen on the Excel file over time will be able to be reloaded without any problem into Tableau.

Tableau Revit

 

All our dimensions and measures are now exactly how we want them for visualizing. From here you can just drag and drop into the worksheet to visualize the data you want. Here are a couple of examples:

 

Tableau Revit

 

Example 1: This is simply a tabular view listing each of the Revit elements that exist with a particular name. You see Revit’s internal GUID as well as the square footage of that room. We are also formatting the cell to conditionally color from smallest to biggest area.

 

Revit tableu by levelExample 2: This is a bubble chart that is meant to visualize proportions between the different space types. In this case we are not viewing every instance of a room type but instead are comparing an aggregation of all the Rooms with a particular use against another. The number of bubbles represent that number of ‘Uses’ represented in the model, and the size of the bubble represents the aggregate area of all the Rooms in that Use category.

 

Some Examples of How BIM Data Can Help Us Make Better Decisions

This is a presentation I gave at last year’s AIA TAP at CIFE. It was a short presentation on some of the potential benefits of mining BIM data and studying it to help emerge some truths, patterns, etc. about our projects, processes and our teams. I showed some specific examples on how we are using some of this data today to make more informed decisions.

Archiving and search

One of the lowest hanging fruit of all this is the idea that if we can access all the information that is stored in these models, we can literally search through our past projects. Imagine being able to search for what doors or what floor to floor height were used on a building you finished 5 years ago, without having to run down to the archives.

Comparative analysis

It is common for us as industry to use precedents, but most of the time we are limited by relatively obvious attributes in our selection. We use things like typology, project size, client, client type and region as good meta organizers. What if we wanted to take that further? What if we wanted to identify projects by their floor to floor heights? or, what if we wanted to identify projects by the use of some particular piece of mechanical equipment? By that same measure, what could we learn if we could take successful projects and compare them against those considered to be less successful?

Smarter QA/QC

When working with large models and large teams, knowing ‘what is in the model’ can be almost impossible to determine. The rate at which people are adding elements is faster than the speed at which we can check them. Our qa/qc tools are not in line with our speed of production. Running model checks against a set performance standard is another big opportunity here. The information is there, we just need to be able to see it in the right form, and it needs to come to us at the right speed. It needs to keep up with production.