See Application

Inspiration

With the impact of Covid-19, supermarkets and shopping malls had to restrict the customer volume. As a result, it is very normal to observe that many customers lined up at the entrance of the malls which is both time-consuming and also increase the risk of getting infected.

What it does

Our new app ‘See’ which can help customers to know the real-time customer volume of the malls and make an appointment to avoid standing in a long queue.

How we built it

To build this app, we assumed two data resources. The real-time customer flow data comes from the installation of infrared devices in supermarkets. Besides, we also assume that the other half of data comes from the customers who use our App to make an appointment. Multiple data resources significantly improved the accuracy of our model prediction. Then, we use Poison distribution to simulate the number of customers and applied Gaussian process to make the prediction.

Challenges we ran into

Due to the time constrain, we are unable to link the system and UI together.

Accomplishments that we're proud of

We clearly figured out a very tricky social issue and built a system to solve the problem! Our system can reach a maximum overall average prediction accuracy rate to 95%!