LEEP // NASA

•

Sleep Shift Scheduling Tool

Sleep loss and fatigue may lead to reduced performance and an increased risk to safety during many activities, including spaceflight. Your challenge is to develop an operational sleep shift scheduling tool that provides autonomous customization of a schedule for sleep, exercise, and nutrition to manage fatigue.

Summary

A good night's sleep is essential for our health and keeping us alert during the day. When astronauts are in space, they are in an inhospitable environment, and a simple activity like sleeping becomes a challenge. However many people on Earth experience sleep difficulties as well. Thus inspiring us to provide optimized sleep for everyone. Our Leep app comes to help with this problem. We created a schedule sleep schedule tool, integrating food, exercise, and a module that helps you to correctly pass through sleep states. Thus, with the use of the application by the astronauts, we can improve and teach the APP to take the best actions, and in the future make it available to people on Earth.

How We Addressed This Challenge

Everyone knows the importance of sleep and how a bad sleep can disrupt our circadian rhythm. This problem affects especially the crewmember in space activities and can lead to significant fatigue-induced errors and long-term sleep loss. We can see the effects of a bad sleep in everyone on Earth and the ways astronauts improve sleep may help people around the world.

The main objective is to develop a solution that takes care of your sleep shift scheduling and improves your sleep. To ensure this, we created a platform that contributes to sleep improvement while you are sleeping, through the issuance of Binaural Beats which is a technique of combining two slightly different sound frequencies to create the perception of a single new frequency tone.

Leep APP is an intelligent assistant that creates the best sleep scale so that you can have optimal sleep. We created the Leep module, where you can add an event that changes your circadian rhythm. We call it the Leep Event. You have two options: the Travel Leep, which is about adding a flight and through the information available it is possible to organize the sleep schedule preparing for that trip. Another option is the Sleep Leep, that is able to change your bedtime or wake time or both. That way, with each new Leep Event or edition of an existing one, our APP will reorganize and suggest a sleep schedule.

Our charismatic avatar, the Little Leep, gives you a personalized and automated experience building your sleep shift scheduling, seeking the best for your circadian rhythm. He is a responsive assistant that expresses the state of its user in itself, through 5 states of emotion to demonstrate how unbalanced your circadian rhythm is, going from healthy (very happy) to tired (very sad). In this way, he shows the user a different way of dealing with sleep and circadian rhythm, making people more aware of how their activities affect their sleep and how their sleep affects their activities.

Leep is divided into cycles and has 3 stages, there are studies that show that these stages can be induced with sound frequencies. People with sleep problems fluctuate between their lightest stages and generally have little time for deep sleep, causing them to wake up without feeling rested. The Machine Learning intelligence of Leep, through a real time collection of data by an non-invasive electrode inserted into a headset, analyzes your sleep quality and maintains your correct sleep frequency for the best possible rest, generating the best sound frequency that can correct your sleep stage and ensures that you pass through all the sleep stages correctly. This functionality is available in the Go to Bed function, in our APP.

In the image below, we visualize the three sleep states (Alpha, Theta and Delta). To have a good sleep you must pass through Alpha, Theta and arrive in Delta where deep sleep occurs and makes you relax. Then you will leave Delta and go to Alpha to wake up. However, we are not always able to follow this path. So our application through the non-invasive electrode, is able to know in which state you are and help you to stay on the correct path through the reproduction of low frequency sounds.

The nutrition and exercise modules will contribute to better sleep and adaptation of each individual's biological clock. These two modules will integrate with applications already on the market. Our future plan is to synchronize with good nutrition and exercise APPs so that the Leep is complete and can provide a fully integrated and personalized agenda. We also plan to provide synchronization for other devices such as Mi Band and Apple watch, in addition to our headset, which is a simpler and cheaper device.

So we hope to get with our APP, a tool that is used by astronauts and with the help of NASA to applaud and train the intelligence of the Leep so that in the future it will be available to anyone in the universe.

How We Developed This Project

We used material from various articles and projects related to the study of sleep in the Nasa dataset [1] to fundamentalize the architecture of our system. In addition, we also combine with different state of the art studies [2] [3] [4], ensuring a solid foundation.

Our implementation is based on a CNN model [6] that consists of four convolutional layers and two max-pooling layers. Each convolutional layer performs three operations sequentially: 1D-convolution with its filters, batch normalization, and applying the rectified linear unit (ReLU) activation. Each pooling layer downsamples inputs using max operation. Then, we use two layers of bidirectional-LSTMs to learn temporal information such as stage transition rules which sleep experts use to determine the next possible sleep stages based on the previous stages.

How We Used Space Agency Data in This Project

Because we are waiting for permission to access the NASA biosignal dataset [5], we used external electroencephalogram databases so that we can train and validate our neural network.