Bodyguard

To protect intellectual property, I have omitted descriptions of the workflow and details about the research and design process.

Overview 📘

One of my key projects as a UX Designer on the Alexa Human Effects (HFX) team was “Bodyguard”, a feature designed to improve user safety by detecting environmental sounds while using noise-canceling earbuds. This feature was born out of customer research suggesting its usefulness and consumer desire. I synthesized research data to advocate for this feature, designed its user experience, and iterated based on stakeholder feedback.

The Problem 💢

When users wear earbuds—especially with active noise cancellation (ANC)—they become less aware of their surroundings, which can pose safety risks. The challenge was to design a feature that could make users situationally aware.

Research 🔍

I synthesized internal research and customer surveys to identify user needs. Key takeaways included:

• Being warned of danger was a highly desired job to be done from technology for both Alexa and non-Alexa users.
• There was a strong demand for hearables to detect and notify users of important environmental sounds.

Defining the Feature 🔉

Based on this research, I narrowed down on designing a feature for Echo Buds that could detect certain sounds - but which sounds? I created a sound priority list based on:

• Survey data on customer preferences for sound detection.
• User scenarios where situational awareness would be most beneficial to their safety.
• Pre-existing sound detection models for other Alexa products that could be leveraged.

A key limitation to chosen sounds was Alexa's inherent processing latency on Echo Buds, which affected real-time sound detection. This meant that some urgent sounds might not be useful for immediate intervention, leading to certain sounds being deprioritized for being less useful in such scenarios.

Design 🛠️

After mapping out user workflows and iterating on the user experience based on research and feedback from the team, I created documentation and audio demos demonstrating the user experience of Echo Buds reacting to long-duration sounds and short-duration sounds.

Depending on the state of the Echo Buds at the time of interruption (music, podcast, calls, nothing), I outlined:

• Whether audio playback should be ducked or paused
• Whether a sound cue (earcon) or text-to-speech notification was sent.

These specifications adhered to Alexa’s voice interruption design guidelines.

Next Steps 🤔

The findings and workflow I created were incorporated into the product roadmap, allowing engineers and product managers to refine the feature further. Follow-ups I recommended included:

• The development team should optimize sensitivity thresholds to balance when an interruption was triggered while minimizing false positives.
• The audio engineering team should design a unique sound cue (earcon) to help users intuitively recognize Bodyguard alerts.
• Consider designing beyond safety-related sounds. The sound list I created prioritized safety. But if we had the opportunity to expand upon the feature, users could find value in detecting other sounds that were not directly related to safety.