Every one of us has certainly a kid they worry about (Own child, sister/brother, niece/nephew...), these children are continuously exposed to inappropriate content, scenes or acts (violence, bullying, fights....) that triggers their fears expressed by their level of stress or heartbeats. Today's children are tommorow's adults, so investing in making a better childhood for them is a long term investing to make a better world tommorow.

What it does

The purpose of this project is to monitor and watch over kids as they develop curiosity for daily content and experiences. By watching our kid's mental health peaks we aim to prevent the worst scnearios from happening such as lack of self confidence, mental health issuses or even willingess to make crimes. We will get vital signs data such as stress level and heart rates from a wearable device on the child’s wrist through Terra API in real-time. The data will arrive to the heart of our system : our backend, which will ask our model for a prediction on the child's mental state. It will update the state in the parent's mobile app with a color code (Green: your child is okay, Orange: your child may be in a stressful situation, Red: your child is in highly stressful situation). By then, a notification will be sent instantly to the parents' device to notify them with the level of stress and the mode of control they choose (Shutdown the device, Blur the device's screen or do a parental check directly) to make a retroaction to the situation.

How we built it

With the partner of Terra API, we used the following : Methodes : -Machine Learning. -Predictive Modeling. -Data science. Technologies : -Python, Flask. -Pandas, jupyter. -Flutter. -HTML, CSS, JavaScript. -Node js & Express.

Challenges we ran into

We faced some problems when trying to access data through Terra Api, the server was down at some points, and it took us a considerable time to configure and understand the API’s concept. We didn’t have access to a wearable device to simulate the data flow, from a kid's wrist to our server. So testing our model was made with static data since we couldn't generate fake data related to the kid's stress level and heartbeats.

Accomplishments that we're proud of

We are proud of making a 36 hours sprint with only 5hours of rest. We are proud of having learned so many new thechnologies and methodes and practice them in a project that solves a real world problem. We are proud of designing a solution that is humanitarian and will make a big impact on communities future.

What we learned

Discovering Terra's API. A better comprehension of Machine Learning, Predictive Modeling, Data science.

What's next for KidErra

As a perspective we want to exploit our solution's model to extend it to everyday use, because the disturbance could come from various factor in the kid’s environment.

Built With

Share this project: