Helping parents feel safer through intelligent alerts and personalized sleep insights
Hubble Connected • 2022 - 2023 • New feature development
Background
In 2023, Hubble launched Hubble AI, a suite of next-generation AI-powered features built into its smart cameras. Unlike traditional monitors that only offer passive video feeds, Hubble AI features are designed to deliver real-time intelligence and personalized insights supporting safer & healthier sleep routines and greater peace of mind for parents.
My Role
I led the UX design for Hubble AI features, working closely with product managers and the in-house machine learning team to shape these experiences from concept to launch.
My responsibilities included:
Designing the end-to-end experience for both safety alerts and sleep insights
Translating AI logic into emotionally supportive, user-friendly UX
Prototyping and validating UI patterns for notifications, dashboards, and data visualization
Handling edge cases and building user trust, especially in situations where AI might send false alerts.
The Challenge
From user interviews and support ticket analysis, we uncovered two key challenges:
Sleep monitoring was exhausting and manual:
Even with video monitors, many parents felt the need to stay constantly alert, manually track their baby’s sleep routine, and often ended up sacrificing their own rest in the process. At the same time, they expressed a strong desire for greater visibility into their baby’s sleep patterns to build healthy, age-appropriate sleep habits.
Infants' Rollover phase creates anxiety:
Another key area of concern was the rollover phase, a critical developmental milestone where babies begin learning to roll over independently. During this transitional stage, many babies don’t yet have the strength to roll back, leaving them stuck in awkward or potentially unsafe positions. This created both physical risk and emotional stress for caregivers.
Together, these insights helped us frame a simple but powerful design goal:
Help parents rest easier by making baby safety and sleep understandable, visible, and proactive through AI.
Key UX solutions
A. Covered Face Detection Alerts
Clear & Contextual Alerts
In the early stages of this feature, as the models were still evolving, some alerts resulted in false alarms. To address this, we crafted each message using calm, parent-friendly language such as “Your baby may not be on their back” to gently inform without causing panic.
Each alert was also paired with a still image from the camera, allowing parents to quickly assess the situation before taking action. This approach helped build trust and ensured that alerts felt helpful rather than stressful.
Trust through clarity
To build trust in the feature, we included a visual guide showing which sleep positions are safe versus which ones trigger alerts, making the system’s behavior feel transparent and reassuring.
B. Sleep Insights
Quick access to sleep insights
Parents are guided with age-based sleep targets (e.g., 16 hours for infants), helping them stay aligned with pediatric recommendations.
The app shows how much sleep the baby has completed so far (e.g., 8h 15m), reinforcing daily progress and long-term healthy sleep development.
Why this matters:
A consistent sleep target supports both the baby’s cognitive development and the parent’s ability to build effective routines.
Sleep Quality & Positions insights
The app visually breaks down how long the baby slept on their back, belly, or side, using familiar.
If the baby sleeps mostly on their belly, a recommended safe sleep position it’s clearly highlighted to reassure caregivers.
It offers subtle, helpful cues about safe sleep so parents stay informed without needing to monitor every movement.
AI calculates sleep quality based on disruptions and movement patterns, presenting it as a simple percentage (e.g., 82% – Excellent) while tracking rollovers to give parents comprehensive insight into their baby's restlessness during sleep.
Sleep Sessions Timeline
All sleep sessions, whether night-time or naps are recorded and displayed on a consolidated timeline.
Parents can view patterns across the day and night, helping them adjust nap schedules and understand sleep cycles over time.
Impact
Learnings
Trust is built through clarity, not complexity:
In emotionally sensitive spaces like infant care, users don’t want more data; they want clear, contextual guidance. Designing with empathy was as important as technical accuracy.
Transparency drives confidence in AI:
Educating users on what triggers alerts and showing safe vs. unsafe sleep positions made the system feel more predictable and trustworthy, even when the AI wasn’t perfect.
Calm language + visual cues = reassurance
Alert messaging had to be direct, but never alarming. Pairing notifications with friendly visuals helped caregivers understand what happened and feel in control.
Huge relief as a parent
The AI features here are a class apart. It can detect rollover and covered face scenarios accurately, which is a huge relief as a parent. I love that it doesn't send constant false alarms.
AI Vision made it super simple
As a first-time parent, I was nervous about setting up a smart monitor, but AI Vision made it super simple... Bonus: I can track sleep patterns through Hubble AI. Love it!
Helped me understand baby's
sleep pattern
Sleep insights have helped me understand my baby’s sleep patterns better, & the sleep trainer feature is a great bonus for developing healthier sleep habits.
Game‑changer for safety
The AI alerts me instantly if it detects anything unusual, like my baby's face being covered, which is a total game‑changer for safety.
Back to top
