Poli

Key Research Questions:

• How can smart home systems be managed democratically and transparently by multiple users?
• How can ownership of devices and events be defined, especially when owners are unavailable?
• How can the system balance collaboration, privacy, and personalization?

I analyzed popular smart home apps (SmartThings, Google Home, Apple Home, etc.) and found:

• Strengths: User-friendly interfaces, device grouping, and energy monitoring.
• Weaknesses: Limited multi-user features, poor conflict resolution, and inadequate guest management.

• Semi-Structured Interviews: Conducted with smart home users and IoT experts to uncover pain points like latency, lack of haptic feedback, and the need for guest management.
• Surveys: Two surveys targeted shared household dynamics and smart home usage:
  • Survey 1 (Household Dynamics): Highlighted collaboration challenges (e.g., scheduling chores, resolving temperature preferences) and the potential for technology to aid multi-user coordination.
  • Survey 2 (Smart Home Users): Revealed that users often share devices but face issues like lack of activity logging and child safety controls.
• Key Insights:
  • Users value transparency but need privacy options.
  • Conflicts arise from instant changes (e.g., lights) more than gradual ones (e.g., thermostats).
  • Users desire features like activity logs, guest modes, and voting mechanisms for decisions.

Using Trello to categorize IoT UX aspects (e.g., UI types, interconnectedness, decision-making models), I narrowed the focus to multi-user smart home scenarios in shared households. I brainstormed solutions addressing:

• Usage Conflicts: Alerts for recent changes to prevent overrides.
• Transparency: Activity logs and user presence indicators.
• Ownership: Clear roles (owners, sharers, guests).
• Collaboration: Voting and suggestion mechanisms.

I created low-fidelity wireframes to visualize key interactions:

• Conflict Alerts: Notify users of recent device changes with options to confirm or suggest alternatives.
• Activity History: Quick peeks at recent users and detailed logs.
• User Management: Role assignments (e.g., owner, guest) and privacy settings.
• Suggestions: Voting for proposed changes to resolve conflicts.

• Focused on core multi-user features (user management, device control, rule-setting).
• Used realistic input/output methods to test concepts without polished visuals.
• Tested with users to validate hypotheses about collaboration and transparency.

• Refined based on first-round feedback:
  • Simplified user roles (e.g., renamed “restricted user” to “guest”).
  • Enhanced conflict explanations with context-specific details.
  • Improved privacy settings accessibility and clarity.
  • Added a “Let Know” feature for household task communication (e.g., notifying roommates about completed chores).
• Polished UI for consistency and cross-platform adaptability (Android/iOS).

• Tasks: Adding users, managing devices, setting rules, and resolving conflicts.
• Findings: Users struggled with unclear labels (e.g., “restricted user”), complex rule-setting, and accessing privacy settings.
• Metrics: Rated tasks on ease (1-7) and time efficiency; hypothesis validation showed moderate support for multi-user features.

• Updates: Added task communication feature, simplified labels, and improved privacy settings flow.
• Findings: Improved ease and time efficiency scores; users appreciated clearer conflict explanations and the “Let Know” feature, though privacy settings placement remained challenging.
• Net Promoter Score (NPS): Added to measure overall product likeability, showing positive user reception.

A senior IoT researcher at Georgia Tech conducted a cognitive walkthrough, suggesting:

• Larger text for privacy settings descriptions.
• More flexible privacy options (e.g., per-person visibility).
• Replacing “owner” with “manager” to avoid confusion.
• Simplifying voting mechanisms for less social awkwardness.

Poli’s key features address multi-user challenges:

1. Guest Mode: Allows temporary access for guests with customizable permissions.
2. Conflict Resolution: Warns users of potential conflicts with voting or override options.
3. Privacy Settings: Flexible controls for location and activity visibility.
4. Ownership Clarity: Defines “managers” and “sharers” for devices and rooms.
5. Quick Peek: Displays recent device users to prevent unintended changes.
6. Double Check: Alerts users if a device was recently used, with options to confirm or message others.
7. Suggestions: Enables democratic or consensus-based voting for changes.
8. Messages & Let Know: Facilitates household communication for tasks like chore coordination.

Based on research and testing, I derived seven principles for multi-user smart home systems:

1. Clear Definitions: Use simple, non-technical terms (e.g., “manager” instead of “owner”).
2. Guide, Don’t Restrict: Facilitate collaboration without overriding user autonomy.
3. Explain Conflicts: Provide specific reasons for conflicts to aid resolution.
4. Balance Transparency and Privacy: Ensure visibility with flexible privacy options.
5. People-Centric Rules: Base rule-setting on users, not devices, for intuitiveness.
6. Integrate Household Management: Optionally include task coordination, ideally tied to devices.
7. Minimize Social Awkwardness: Avoid overly rigid systems that hinder natural communication.

• Children’s Roles: Explore restricted roles for younger users.
• Enhanced Privacy: Offer anonymized presence indicators (e.g., number of people in a room).
• Recording Devices: Address privacy concerns for cameras and microphones in shared spaces.