The University of Arizona’s Path to Enhanced User Experience Through a Custom GIS Routing Tool
Launching an Award-Winning Product
Early August 2023, just weeks prior to launching a custom pedestrian and bicycle routing system for campus, the UArizona Enterprise GIS (EGIS) team (pictured below) was honored to win the Ellington Innovation award for its related presentation at the Campus Facilities Technology Association (CFTA) 2023 Annual Conference in St. Louis.
CFTA 2023 Ellington Innovation Award Winner
"Are We There Yet? The University of Arizona's Custom Pedestrian Routing Network"
The CFTA Ellington award was a great lead-in to the formal product launch, culminating several years of hard work. User and stakeholder needs drove the functional and UI/UX design process, including factors such as:
- Initial launch was planned within UArrive, UArizona’s parking and transportation map application, so the routing tool was designed with the “user-on-the-go” in mind (mobile first, smaller device dimensions).
- Achieving a look cohesive with UArrive’s application styling and UI
- Enhanced user experience by utilizing more accurate, detailed, and up-to-date UArizona curated GIS data.
- Ability to apply custom route filters, such as for accessibility
- Responsiveness to real world conditions, such as construction fences and events.
My name is Grant McCormick and I’m the Director for Enterprise GIS (EGIS) at the University of Arizona, located in Tucson, Arizona. In this post I’ll highlight the excellent work my amazing team undertook in implementing the technical process, along with some lessons learned and next steps.
Building Custom Routing Technology
Going from raw data to a final product included much trial and error but included these overall steps: data development, creating the network dataset and custom locators, publishing services, and custom widget development for user workflows.
Diagram of Custom Routing System Development Workflow
Data Preparation
Considerable data preparation occurred, starting with creating line features classes representing walking and biking routes. We used Arcade to populate a Name field based on other attribute fields (saving time and reducing errors), and other feature classes (sidewalks, building entrances, building outlines, parking lots) informed editing of individual route links.
Creation of the Routing Network
Route data was then integrated into a Network Dataset. This began by adding the data to a feature dataset, then setting a series of properties such as data sources, cost parameters and restrictions for each travel mode.
With network properties set, the last step was to build the network (a function within ArcGIS Pro) so that it can analytically solve routes. The tool was then tested using ArcGIS Pro to ensure correct results.
Creation of Custom Routing Locators
Custom locators were created to enhance user experience with specific route start and end points. Locators provide locations users choose from when searching for starting and ending points. Our custom locator merges three polygon feature classes (buildings, parking lots, open spaces) along with building entrance points, and was subsequently combined with the Esri world locator to create a composite locator covering geographic locations beyond the extent of ours. A locator for accessible routes was created as well.
Publishing Routing Services
After creating the custom network dataset and locators, we published them as services to ArcGIS Enterprise for use in online applications. This included adding the network dataset as a route layer to ArcMap and enabling network analysis when publishing. To publish locators, we used ArcGIS Pro, selecting the share locator option under the Share Ribbon.
Custom Widget Development for End User Workflows
The Esri Javascript API was used to build a custom routing widget (refined user sequences, design of UI elements) to optimize end-user workflows. When entering a location in the search widget or clicking a location on the map, this information is consumed by the custom locator to match addresses. Solved routes are then added to the map using custom graphics.
Custom Routing Widget
A Walking Route Solved Same Route Solved with Accessibility Filter
Lessons Learned and the Route Ahead
The project development process involved substantial trial and error and adapting technology components to fit custom needs. Among the many lessons learned were:
- Intersecting network lines are automatically split at their junction, causing attributes provided through domains to be lost due to default domain split policy. Split policies were changed from Default to Duplicate so that lines resulting from a split preserve the original line’s attributes.
- Elevation parameters were not enabled in creating the initial network dataset, resulting in route errors with underpasses and underground plazas. For example, an underpass sidewalk line intersected with a sidewalk line along the street above, so the user would be incorrectly switched from underground to ground level. The network dataset was subsequently adjusted to respect elevation changes.
Vertically Separated Route Segments
- We found no publishing issues with only one travel mode. However, as additional travel modes were added, directions for the full routes did not display. We found a work-around for publishing the network dataset by using ArcMap instead of ArcGIS Pro.
Building on the successful product launch we envision several future routing tool enhancements, such as:
- An immersive 3D user experience for viewing routes.
- Incorporate route preference (such as shaded routes, routes near street lights).
- Include a street network for use with the driving transportation mode.
- Extend benefits of the routing system by deploying in other web applications.
- Take the routing system indoors.
Immersive 3D Routing Prototype
If you have any questions, please comment below or feel free to reach out to me at grantmc@arizona.edu.