Spartan

Inspiration

Every morning, Tim's alarm goes off. He snoozes it, then snoozes it again, and then scrolls on YouTube shorts for a while before finally getting out of bed. He looses over 30 minutes a day, everyday, for nothing. Tim, like most of his peers, suffers from laziness. Well Spartan is about to change this. Studies show that light, sound and physical exercise, all decrease melatonin and adenosine levels and increase cortisol levels in our bodies. This productivity tool combines all three and forces the user to wake up as quick as possible.

What it does

The user is able to create customizable alarms that include music, motivation speeches and light shows. But most importantly it is connected to a camera that tracks the user's movement and ensures that they perform the required physical exercises. Until then, there is no other way to turn off the alarm. Indeed, no room is left for laziness with Spartan. Customizable features include music and speech choice and volume, light color and brightness, alarm time, and most impressively, routine exercise.

How we built it

We are running our code on an orange pi, with an OV13850 camera, a 512 SSD and a wifi chip, running ubuntu 22.04.

-> The bluetooth LEDs are from Govee, which has an easy-to-use python api to control them.

-> The tts is implemented through *** .

-> We connected aux speakers and used the drivers provided by the orange pi's ubuntu image to output music to them.

-> The front-end UI is implemented through PyQT.

-> We used a repo https://github.com/aryanvij02/PushUpCounter for the computer vision push up counter.

Challenges we ran into

First, the Wi-Fi was working but the Bluetooth was not. The Bluetooth card could not capture the signal from the Bluetooth led strip. Luckily, a last minute savior brought an antenna for the network chip of the orange pie and the Bluetooth started working. However, the Wi-Fi then stopped working. Moreover, when looking for a text-to-speech (TTS) API for Python, all the common ones were terribly robot sounding, so we had to create an IBM cloud services account and use their TTS service. This one was trained using deep-learning and sounded very human-like.

Accomplishments that we're proud of

We fully made the UI using Python and CSS, like true hackers. The video edit of the design process turned out very nice, everyone in the room, including the security guard that participated in it, enjoyed it.

What we learned

Through a lot of troubleshooting and debugging, we all developed our problem-solving abilities and lateral-thinking. Indeed a lot trouble was encountered when trying to get the Wi-Fi and the Bluetooth working.

What's next for Spartan

The future of Spartan lies in sleep tracking. With the ability to connect the application to smart watches, we would monitor sleep and set the alarm in function of sleep cycles instead of time durations in order to optimize sleep efficiency.