/Hackathon-SingaporeAirlines

This repository contains submission by Team Blitzkrieg to Singapore Airlines AppChallenge 2018

Primary LanguageCSSMIT LicenseMIT

“SnakeEye” is an AI-based system integration platform that uses Object detection & Deep learning techniques to detect & classify F&B into three types Untouched Leftovers Consumed

Our algorithm will then count the number of times each category has appeared and correspondingly update it in the database.

Sources - Video demo - https://youtu.be/R-N4GLBP7N8
Presentation - https://bit.ly/2zGEklL
Dashboard - https://bit.ly/2P11CI2


Current Scenario:

Presently, there is lot of manual intervention involved in terms of counting the wasted meals. Then the numbers are entered manually in a spreadsheet. Wastage(%) is then computed based on the capacity of the flight.

Proposed solution:

Having a system in place that automatically counts & classifies meals Store the count in the database along with other details of the flight from the API ( Capacity, Number of people, Gender, Age, Diet & so on ) Dashboarding everything in a single place for management to track.


Core Technology:

  1. Deep learning: R-CNN implemented using Python & Tensorflow

  2. API’s Provided by SIA: /flightschedule, /flight/passenger, /equipment/loadplan - Using Postman collection to fetch data into the database.

  3. Dashboarding: Node.js, D3.js and MongoDB

  4. Predictive Modeling: Python, Scikit-Learn, Numpy & Pandas