Inspiration

With more people stuck at home due to COVID-19 pandemic, there has been a surge in demand for food delivery services. Due to this, more and more restaurants are offering home delivery/curbside pickup options. With multiple retail options, restaurants are finding it hard to manage their services. This has also caused cases of inconsistency in what people see online and what they receive as deliveries

What it does

To solve the problem of surge in online orders, we have designed a state-of-the-art “SurgeBalance” algorithm that uses ML and Data Science to make sure restaurants are never under the nerve. Our app uses technologies like Augmented Reality to provide consumers a ground-truth of their orders and portion-sizes - thus creating a seamless transition from Brick & Mortar to virtually at home. To top this all, we have designed a data analytics dashboard for restaurants to manage and run their businesses smoothly. All built using NCR’s rich APIs!

We define concept of “busyness” of a branch and quantify it using our algorithm:

  • Multi-Objective Minimization Problem - distance of consumer from branches, and the busyness of a branch Surge Balancing Algorithm:
  • Using available staff, number of active orders, difficulty of preparation
  • “Complexity” of menu item - function of preparation time and elapsed time since order was placed
  • “Busyness” is calculated by a regressive approach based on staff, complexity of all orders
  • Final choice of optimal branch to send order to is the minimum of weighted sums of driving time to customer and busyness

We have also created a loyalty program for our returning customers to prevent them from rerouting to new locations over and over again. The greater the loyalty points, the more of getting lower weight times. This provides an incentive for a loyal user base.

And we are the most "social" delivery application to date - you can create a virtual menu with your friends, split bills evenly, and enjoy the fun of going "dutch" while being at home, even in different states!

Our first Figma prototype designed using NCR's design guidelines: https://www.figma.com/proto/5P5AFAfP3NbJjztrDS37k8/HackGT?node-id=2%3A2&viewport=180%2C331%2C0.35989025235176086&scaling=scale-down

Our React Native + Django app for iOS/Android: https://github.com/gursimransingh93/crust-hackgt7

What's next for Crust

We have built a solution that can be scaled easily with current systems and uses a mobile-first approach to solve a lot of problems that NCR is facing right now. An interesting approach: we plan on placing this as a white-labeled API-as-a-Service that significantly improves and extends/builds upon DoorDash’s white-label Drive service.

Share this project:

Updates