Data analysis for government

UrbanLogiq

The silo problem with government data

City departments currently lack visibility into the data of other departments, limiting their ability to make well-informed decisions for their operations. To access relevant data, they must navigate through a multitude of Excel sheets, PDF files, and disparate legacy systems within their own departments and across different jurisdictions. The problem of data silos is a common challenge faced by cities, counties, regional governments, and states. Our aim is to address this issue by providing a unified platform that empowers public servants, such as Traffic and Transportation Engineers, Economic Development Officials, and City Planners, to make data-driven decisions more effectively and efficiently.

The design process was initiated on October 2019. Launched on June 2020
Product designer overseeing the end-to-end design process, and establishing the design system
Workshops, User research, Information Architecture, UX & UI, Prototyping

WHAT DO WE WANT TO ACHIEVE?

A platform for aggregating data from various sources, manipulate this data into metrics and insights to help cities make better decisions in traffic operations, economic development, and planning.

UNDERSTANDING THE PROBLEM

How do users work with data and what they want from it?

Existent Data

Surveys and interviews made with clients in the past

Benchmark

Competitors : GIS, reports, traffic solutions, etc.

Interviews

Observation of current workflows

A COUPLE OF INSIGHTS

Discovery and monitoring mode

“ I don’t have a specific need to keep checking. I want to layer/combine the data accordingly to my need in the moment. Sometimes I have to help investors to find the best spot in the city for their business.”

“I’m responsible for analyzing the transit conditions and solving problems regard to it. The City is building a new train line and I need to “predict” how the construction will impact the traffic.”

Problem statement
& Design Goals

It is difficult to tease out correlations between data in government because of the siloed nature of the data.

The solution will provide UL clients with an advanced geospatial filtering tooling, the Explore tool, where they can perform the tasks below :

1. Ability to search for data attributes from different datasets, layer the data in a visualization map, and filter by a time window.

2. Ability to combine search criteria from multiple datasets and geofence areas corresponding to the selected criteria

3. Ability to select an area and display a summary of the data on a report with options to dig deeper and change data displayed on graphs.

Challenges

One of the primary challenges in the project revolved around defining the information architecture and user flow for exploring and uncovering correlations within datasets. To tackle this challenge, I adopted a similar approach to designing a search system for textual data. The first step was to understand which fields were indexed to define advanced search filter options.

To accomplish this, I conducted a thorough analysis of a comprehensive sample of datasets, aiming to identify their common structure and discover the various data types they contained. This analysis allowed me to gain a deeper understanding of how queries could be constructed and what types of selectors users would require when interacting with specific data types within a dataset.

Search flow

The search flow involves creating layers and applying filters to combine multiple attributes within a dataset for a more specific query. Users can narrow down their search by selecting criteria such as income range and concentration of renters. They can also join data from different datasets using operators, allowing them to search for specific scenarios like areas with aging populations near hospitals or without hospitals. The filter options vary based on the data type of the selected attribute.

DESIGN PROCESS

The outcome

We opted for a clean and minimal design, prioritizing high contrast between colors to ensure clear differentiation between data and other UI elements. To accommodate user preferences, we developed both light and dark themes for the interface.