Discover: Empathize & Define
Research Goals
I aimed to investigate how data scientists could benefit from a highly visual and geospatial tool that enhances their ability to identify potential wildfire and PSPS risks.
Research Methods
To achieve this goal, I connected with stakeholders and users weekly to understand their current processes and identify opportunities for improvement.
The research aimed to achieve the following objectives:
1. To gain insight into how current users identify potential wildfire and PSPS risks and their existing processes.
2. To determine the value of a geospatial web tool in enhancing decision-making abilities during emergencies such as wildfires and PSPS events.
3. To understand the process of identifying potential PSPS threats and explore how machine learning can be utilized to prevent them.
4. To explore the trade-off between providing a snapshot view of current data and the need for in-depth analysis.
User Interviews, Workshops & Exploration
In addition to meeting with stakeholders and internal users weekly, I conducted workshops with 10-20 utility employees who work with the data that determines emergency events daily. The workshops aimed to define various types of users and identify current setbacks and opportunities using the "How Might We" and "5W +1 H" techniques. Furthermore, the workshops helped to define the Golden Path for the project. I also conducted various sessions for prototype exploration and review to ensure that the design met the end users' needs.
Key Findings
Our research and discussions revealed the following key findings that needed to be addressed to achieve our goals of strengthening SoCal's resilience against wildfires and PSPS events:
1. The current tool lacked a visual interface and required users to utilize various tools to complete their work, which made the process time-consuming and less efficient.
2. The update intervals needed to be more consistent, which affected the data's accuracy and relevance.
3. The data models were not user-friendly and required a high level of expertise to digest, which limited the tool's accessibility to a broader range of users.
4. The current tool could not integrate with external apps, which limited its functionality and flexibility. The risk scenario management needed to be automated, as it left room for human error and increased the risk of missing potential risks.
Overall, these findings highlighted several areas for improvement that could be addressed by creating a more user-friendly and visual tool, adding external features such as maps/dashboards, automating risk scenario management, and ensuring consistent and up-to-date data.
