GateInventory

main screen
menu selection by flight
inventory
Result of the analysis
List of variables for a flight

How we built it

We created a computer vision system using convolutional neural networks to detect and classify items. The data is processed through a pipeline for analytics and reports, and everything can be monitored through a Flask-based web dashboard.

Challenges we ran into

It was hard to get accurate object recognition with different lighting and angles specially since it was our first dive into ML and vision. We also had to make sure the system worked well with limited computing power and combine several data sources into one clear platform.

Accomplishments we’re proud of

We built a working prototype that recognizes inventory items with good accuracy, handles real-time data from multiple inputs, and includes a simple control panel to visualize complex information.

What we learned

We learned how to train and optimize computer vision models for specific environments, how we could use AIs to simplify some tasks and pipelines, and understand the challenges of airline catering and logistics.

What’s next

We plan to expand the item recognition system to include more catering products and add predictive analytics to forecast consumption and improve efficiency.