Eveil

Framing the Problem

What is the problem you are trying to solve? Who does it affect?

Alarm clocks have existed for over 150 years and every innovation in that time has optimized the signal — the sound, the light, the timing. However, nobody has closed the loop on whether the person actually gets out of bed. For millions of people, particularly those with ADHD and high sleep inertia, the obstacle is not awareness that it is time to wake up. They know. The problem is the physical inability to initiate leaving the bed even when fully conscious of the need. This is not laziness, rather it is a genuine neurological challenge that no alarm product has ever directly addressed. The snooze button exists because the industry accepted defeat.

Within our team, Kush proposed the idea as he personally saw the problem as one he faced. Being someone with ADHD he constantly finds it a challenge to fulfill his goals while being challenged by his mind at times. Even struggling to make it to this hackathon for the opening ceremony, we saw how deep this problem was.

Idea Explanation

What is your idea? How does it fix the problem?

Éveil is a hardware-software blend that closes the loop on waking up. The device sits under the mattress and uses four force-sensitive resistors to monitor whether the user is in bed. The paired iOS app runs a smart alarm that fires at the right moment before it’s time to wake up and get ready for a calendar event. The alarm can only be dismissed by physically leaving the bed, which is enabled by the sensors comparing constant pulls of the pressure at any given moment compared to the onboarding calibration.

There is no snooze button. There is no tap-to-dismiss. The alarm stops when the sensors confirm the bed is empty. As a safety measure, the alarm will sound again if the user gets in bed within a short time frame after the alarm sounds.

Implementation

How do all the pieces fit together? Does your frontend make requests to your backend? Where does your database fit in?

The hardware device contains four FSR sensors wired through a voltage divider circuit to an ESP32 microcontroller. A 3D printed force concentrator plate sits above the sensors, focusing distributed body weight into readable pressure signals. The ESP32 communicates with the iOS app over BLE.

The iOS app is built in SwiftUI and handles calibration, sleep tracking, the alarm engine, and calendar integration. Calibration captures an empty and occupied baseline for each sensor zone. The alarm engine uses AlarmKit to surface a full-screen system interrupt that bypasses Do Not Disturb — the only action available is opening the app. The alarm fires within a window calculated from the user's first calendar event minus their preparation buffer minus a learned exit time buffer, targeting a predicted 90-minute sleep cycle trough. Given the health data we also made a conscious choice to prioritize privacy. All processing happens on device, with BLE as the only communication layer between hardware and app.

Challenges

What did you struggle with? How did you overcome it?

The hardest problem was adjusting to the new hackathon environment, as 3 of the 4 members had never been to one. The process of quickly transitioning from idea formation to execution was a challenge, but taught us a lot. Most importantly, our biggest lesson was the value of problem solving: running into an issue and being able to troubleshoot efficiently. In the physical sphere, learning to 3D print, connecting the phone to hardware via bluetooth, etc. was a learning process, but one we’re glad we took. Among the smaller issues, we had to rethink the wake formula entirely: shifting from sleep stage detection to a system with a holistic view incorporating waking times and habits. The physical enclosure also required iteration, in order to consider all the components fitting inside and how to attach them without taking a lot of space. We revisited the drawing board a great deal but are glad we stuck through and saw our vision come to life.

Accomplishments

What did you learn? What did you accomplish?

We built a working pressure sensing circuit where four FSRs wired through voltage dividers to an ESP32, producing stable 0–4095 ADC allowed for readings in real time. We designed and prototyped a 3D printed enclosure with a force concentrator plate that translates distributed mattress pressure into focused sensor readings. In addition we designed a complete iOS app architecture including a calibration state machine, BLE communication layer using the Nordic UART Service, and an alarm engine built on AlarmKit rather than local notifications (a deliberate choice to ensure the alarm surfaces as a full-screen system interrupt that bypasses Do Not Disturb or getting lost amongst users many notifications). We also worked through a complete product going over everything from the PRD, onboarding flow, wake formula, and an ML roadmap for how the system learns and improves over time.

Next Steps

What are the next steps for your project? How can you improve it?

Our ideal next steps include working to create the same product in more of a production (smaller) formfactor along with trying to learn new skills like soldering to achieve a more refined product. On the software side we hope to fine tune small features to enhance the user experience. We have features planned including a morning brief which is able to gather insights from your email and main communication channels to catch you up as soon as your day starts. We also hope to potentially allow users to do a morning brain dump of their dreams, like a journal, to allow their unique sleep story to grow with Eveil.