Playbook
An AI-integrated decision support system for the future success of space missions.
CLIENT
NASA Ames
Research Center
ROLE
Researcher
Designer
TIMELINE
Capstone Project
8 months
April - December
UXD TEAM
Jilly Li
Shivam Shukla
Nhi Tran
SKILLS
UX Research
Prototyping
Usability Testing
TOOLS
Google Forms
Figjam
Figma
Project Overview
Playbook, NASA's mission planning tool, supports human and robotic space missions but relies heavily on manual processes involving numerous individuals. As space exploration advances, mission planning must evolve. With the rise of AI and machine learning, this project aims to design research-backed concepts for an AI-driven mission planning interface in Playbook, tailored for future deep-space exploration.
My Contribution
I was actively involved in the research phase and took the lead in the design process. My responsibilities included conducting a literature review, designing and facilitating user research activities, crafting research questions, interviewing participants, and analyzing data to inform design decisions. I developed and iterated prototypes and also conducted usability testing sessions to refine and enhance features. Additionally, I created marketing materials, including a roll-up banner and a demo video, to effectively showcase our work. As the main point of communication, I ensured effective collaboration and coordination among my teammates.
Video edited and created by Nhi Tran
ASSISTANCE PANEL
The interactive chatbot with AI assistance guide users to navigate complex decisions.
- Assistance Panel
- Suggestions Generator
- Validation Metrics
TIMELINE VISUALIZATION
Simplifies complex information empowering users to make well-informed decisions.
- Guided Timeline with Hints
- Multi-days Overview
- Comparative Timeline View
- Customizable Timeline Editor
REAL-TIME WARNING
Provide a more efficient and non-disruptive solution for crew members.
- Real-time Warning Bar
- Popup Suggestions
PROCESS
Empathize
Empowering Design Decisions Through User-Centered Research
PRELIMINARY RESEARCH
Understand the challenges general people face during the planning and rescheduling process
-
Open-ended, unstructured interviews (n=6): Explored how individuals plan for personal or work-related tasks, uncovering their strategies and identifying any pain points they faced.
-
Group re-planning activity (n=7): Conducted using each group's personal capstone project timeline. Participants were given time and resource constraints for their mid-quarter deliverables and asked to re-plan their entire project under additional constraints. Responses and feedback were collected both manually and through surveys (n=21). This activity simulated autonomous planning in a space mission context.
DATA COLLECTION
Subject Matter Experts (SMEs) Semi-Structured Interview (n=7): Conducted with individuals who have prior experience using Playbook to gather data relevant to the research questions. The interviews were intentionally kept semi-structured and open-ended to thoroughly explore the challenges they face and their expectations, particularly in the context of Playbook, mission planning, and artificial intelligence.
DATA ANALYSIS
After transcribing and familiarizing ourselves with the data, we coded the initial findings by identifying relevant and significant points aligned with our research goals and questions. We then analyzed patterns within the coded data and, through iterative reviews, developed subthemes and overarching themes. This analysis provided qualitative insights that deepened our understanding and guided the development of key themes and future design concepts.
Coded the Initial Data
Identified Initial Themes
Synthesized Themes
RESEARCH KEY TAKEAWAYS
Our primary research objective:
How can we simplify and make the decision-making and mission planning process more efficient for our users through the use of decision support systems?
Simplify and Visualize Complex Information
Allow Users to Explore Multiple Pathways
Ensure Human-Centered and Transparent Workflow
Define
REDEFINED PROBLEM STATEMENT
Suppose there is an AI assisted decision support system designed to simplify and assist the decision making process in contingency planning
AI ASSUMPTIONS
- Integrates and oversees all data within Playbook
- Evaluates and determines parameters like risk, workload, and other key metrics
- Monitors real-time updates (e.g., unexpected weather or missing resources) to adapt plans dynamically
- Utilizes historical plan data as a reference for decision-making
USER TYPES & USER FLOW
Mission Planners
- Need: Simplify complex information and provide assistance during contingency planning
- Challenges: Dealing with a repetitive manual process and managing the uncertainty and complexity of unexpected events
- Goal: Develop and maintain a safe and executable plan for crew members
Crew Members
- Need: Quick and transparent assistance during execution
- Challenges: Working with plans they did not create themselves
- Goal: Successfully complete tasks and activities
Ideate
CONCEPT TESTING
Understanding and Designing Features to Support Autonomous Decision-Making
Using our findings, we brainstormed ideas and consolidated them into three key concepts. We then conducted concept testing (n=2) with direct users to simplify complex workflows, improve transparency, and enhance usability.
Assistance Panel
Provide users with clear, actionable suggestions to explore complex pathways and predict potential options efficiently.
Users Feedback:
- Provide more informative and transparent AI suggestions with detailed explanations to build trust
- Simplify the process by reducing steps
- Present fewer, clearer options with meaningful names
Timeline Visualization
Simplify complex information by visualizing constraints and interdependencies through multiple views.
Users Feedback:
- Enhance clarity and contrast in comparisons
- Include an overview to evaluate future impacts
Real-time Warning
Provide real-time updates and instant user assistance
Users Feedback:
- Minimize disruption from warning bars
- Dropdown menus are inefficient and lack clarity
Evaluate
USABILITY TESTING
Ensuring the design caters to various skill levels
We conducted moderated usability testing (n=11). Our testing prioritized improving the product's flow, with visual design enhancements planned for future iterations if time permits.
- Playbook experience: proficient (n=6) and minimal (n=5)
- Gather insights from both experienced and novice users, ensuring the design caters to various skill levels
Prototype
LO-FI & MID-FI
Hjhj
Acknowledge
LIMITATION
Challenges and Constraints of the Project
- Limited access to the Playbook, which serves as our primary resource
- Restricted access to end-users for direct interaction and feedback
- AI is treated as an assumption, with its capabilities excluded from consideration
ACCOMPLISHMENTS
Self-Initiated Collaboration and User-Centered Solutions
- Proactively initiated and established the project with our client
- Collaborated throughout the research and design process with support from NASA Ames Research Center mentors
- Designed and implemented a prototype that tackled user pain points effectively within existing limitations
- Successfully managed team dynamics and resolved conflicts to achieve project goals
WHAT I'VE LEARNED
Key Takeaways from the Capstone Experience
- Collaborating with NASA Ames Research Center's Playbook team was an invaluable experience
- Improved teamwork skills, including acting as a mediator to maintain a balanced team dynamic
- Developed active listening and clear communication to address ambiguities effectively
- Learned the importance of scoping down and focusing on specific problems to optimize time and resources
WHAT'S NEXT
Some Points We Consider As Future Work
- Evaluate features across various devices like mobile phones and other portable devices
- Investigate in alternative approaches for relevant metrics and their presentation
- Understand AI capabilities and refining design features accordingly
- Conduct usability testing with crews
