Fast Friday

Mockup Screenshot
Resulting Image using a Mask to Attenuate the Traffic along with Entrances/Exits and Customer's Initial Location
Randomly Generated Entrances/Exits
Randomly Seeded Customer Initial Postion

Inspiration

The holiday season is infamous for being a stressful time at the end of the year. As companies look forward to November to go "back-in-black" they will anticipate the largest amount of foot traffic through their stores and offer their best bargains in the shopping calendar. We want to help both customers plan their fastest route, and store managers plan the most lucrative store layout.

This is our most favorite time to shop and we sought to build tools to help alleviate the logistical issues that all of us encounter at the stores this November. The primary motivation is that this improves the retail experience for consumers and that the overall sentiment during Christmas takes on a better sentiment. In the end it is about buying great gifts for your loved ones.

What it does

It helps to provide an efficient path for a customer to buy everything on their Black Friday list so they can have the best route to purchase their holiday presents.

How we built it

We implemented a custom A* graph traversal algorithm, Numpy Methods, Image Thresholding, Signal Attenuation, Image Masking, and made Data Preprocessing functions all in a Jupyter Notebook for Python.

Challenges we ran into

Finding the edges of the image to represent a wall in a store proved to be very difficult! We imputed the signal data along with emphasizing the pixel's size in the image. That helped us to overcome this challenging problem.

Accomplishments that we're proud of

Built custom edge detection methods in our model and practiced data preprocessing techniques. We are glad to hack together our model rationale in such a short amount of time. Usually school projects don't encompass a microcosm of Data Science tasks but we were able to quickly run through Preprocessing, Model Training, and Storytelling tasks.

What we learned

We figured how to leverage and apply Digital Image Processing techniques, how to create specific AI tools to emphasize the contours in the model. Additionally