Throwback Thursday: #HackTheHub18 - A look at AI and Machine Learning

An overview of team #ixdbelfast’s concept ‘FoodPal’

Back in March, myself and a few others from Belfast Art College got together and took part in HackTheHub, an event involving local developers, designers and entrepreneurs to take on a variety of business challenges using AI and Machine Learning.

Brainstorming & Problem Solving

With just hours on the clock we started on paper to get an idea going. Focusing on healthy eating we wanted to create a product that used machine learning to help enforce positive behaviour changes in people’s diets.

The problem with other calorie counting apps is that they take a long time for the user to manually enter all of their food for the day. This becomes tedious and the user gets bored after the first few uses.

With a solid concept, name, and use cases we started wireframing and getting a solid user flow together.

FoodPal

FoodPal is a smart camera app that uses deep learning to track your nutritional habits from your food images.

How does it work?

• FoodPal works by allowing the user to take a picture of their meal and gradually learns their eating habits over time
• The user snapshots their meal and using Google Vision’s image recognition can identify the different foods and estimate the calories and nutritional value
• Over time FoodPal will learn about the user’s food habits and begin to make nutritional suggestions to help encourage positive behaviours
• Vegan/Dairy Free/Gluten Free users can also use FoodPal to detect issues with food in restaurants or shops

Target Market

The FoodPal concept was targeted mainly at millennials who have an active interest in their health and wellbeing.

This group of people in particular are helping drive the demand for mobile health tools along with the growing trend of healthy lifestyles and food choices.

Research also suggests that over half of millennials have an Apple smartphone, and the sheer volume of health app downloads in 2017 shows the growing popularity of fitness trends.

Selected Platform

Technologies

• Would be built in React Native for modern and easy cross platform development
• Uses Google Cloud Vision API to understand the content of an image by encapsulating powerful machine learning models in an easy to use REST API

Basic User Flow

Business Model

FoodPal would run on a freemium business model being free to download initially. A premium version with more features will be then be available for a fee.

Prototype & 2 Minute Pitch

Below is a basic prototype of FoodPal’s flow and how it would function. We gave a 2 minute pitch at the end to explain our idea.

Overall we had a brilliant 2 days and I would like to thank HackTheHub for organising the event, as well as all of the sponsors who made it happen— I look forward to next year!