The ADOT MPD’s (Multimodal Planning Division) journey in search of GIS/LRS data visualization technologies started 6 years ago. The key mission was to produce dashboards that were
- Sustainable – automated with no or minimal coding,
- Performant – response time under 3 seconds, and
- Intuitive – who reads a manual?
By the spring of 2018, however, it was clear that the initial attempt had failed to meet any of the criteria. That summer, the journey resumed when Gistic joined the team under the mandates of leveraging the software licenses acquired and following industry best practices. The team quickly designed this technical strategy – FME as the ETL and automation tool to maintain a SQL Server datamart, which services dashboards designed and published using Tableau.
The software trio delivered! Time for champagne?
But the Map…
Well, the fourth criterion came forth before we could celebrate:
- Great map UI – GIS sells!
For many years, Tableau has been at the apex in the Gartner’s BI Magic Quadrant with its rich functionality and flexible design and implementation options. Tableau’s map support may be adequate for average users, but it can hardly pass muster with GIS professionals. The search circled back to the Esri product line.
As an ArcGIS Enterprise Portal application, Insights enjoys access to any resources ArcGIS Portal has to offer – map services, feature services, geoprocessing services, web maps, as well as non-spatial data. However, Insights lacks the visual design control that is needed to publish dashboards that users expect. After all, Insights was designed as an ad-hoc data exploration tool for GIS analysts and data scientists.
Enter ArcGIS Dashboards, formerly known as Operations Dashboard.
After producing the ADOT structure dashboard using ArcGIS Dashboards in only one day, Karol Francis, who was behind most of the ADOT tableau dashboards, could not conceal her excitement: “ArcGIS Dashboards is so cool! It is really intuitive and delightful to design and develop.”
The superior mapping capabilities come from ArcGIS Dashboards’ ability to leverage Esri web maps, where the symbology, basemap, and attribute popups can be fine-tuned to the likings of a GIS Pro. Filter (called selector) setup, widget (charts and KPIs) placement, and configuration can be done all within the web UI, without the need for separate desktop software licenses!
Dashboard deployment in ArcGIS Dashboards is painless. It is just a matter of making sure the dashboard and all supporting data sources are set to share with the intended audience, whether you are using AGOL or Enterprise Portal.
ArcGIS Dashboards has earned its place here at ADOT MPD, however, it is unlikely to replace Tableau in the foreseeable future. Here is a summary of limitations that the DataViz team has come to realize:
- Table view support. We were able to create a tabular view by embedding HTML in the list widget, however, we could not incorporate any roll-up statistics in the view.
- Calculated fields. Yes, you are able to generate computed fields with Arcade in the beta release. Are you ready to learn yet another scripting language?
- Performance. Compared with a comparable Tableau dashboard, the dashboard using ArcGIS Dashboards lags. The difference is more pronounced when the data size grows larger. The good news, however, is that ArcGIS offers various tuning opportunities for performance improvements. We were able to decrease loading times by 50% by implementing several performance tuning strategies. Stay tuned for a performance optimization white paper, coming soon!
- Widgets. The limited selection of widgets and the limited configuration options of widgets seem to curtail dashboard designers’ expression.
While there are some important limitations when compared with Tableau software, ArcGIS Dashboards excels in the ease of design and deployment as well as in the richness of spatial data representation and interaction. ArcGIS Dashboards has made dashboard creation fun and accessible to a vast GIS community that has long been yearning to use spatial and non-spatial visuals in understanding and presenting data.
This article is based on ADOT MPD white paper on Dashbaording Software Evaluation, prepared by the ADOT DataViz team. Please contact James Meyer at email@example.com if you are interested in more details.