We have all had hard days where we place things and forget where they are. For example, leaving a wallet somewhere in the house and forgetting where it is causing unwanted stress. We are trying to solve the problem of finding your belongings and locating them easily. We are driven by our aim to never let anyone misplace their belongings again.

What it does

It uses the camera on an iPhone and uses it to track the objects around an individual. It also constructs a 3D map of the space around you, so you can pinpoint the misplaced item in a space familiar to you. Conveniently, the app also takes photos of the objects when they were last seen, hence allowing the user to see the missing object exactly.

How we built it

We used technologies like:

  • ARKit
  • CoreML
  • SpringBoot
  • Web GL
  • Web Sockets

Existing technology available was used with standard SDK's to build the app and transmit the data to the server. The web interface was created using a mixture of WebGL and Materialize Styling Library.

Challenges we ran into

Poor documentation on ARKit

It was extremely hard to adapt to non-standard formats of output. For example, the coordinates system used had to be transformed to a consistent format between phone, server and the web client. It required a lot of critical thinking and deduction to understand and normalize the directions and vectors.

React JS didn't work

Integrating WebGL (Three.JS) with React as a component did not go to plan due to our lack of experience writing efficient React code. In the design architecture, we ended up rendering our 3D Plot several times per second. So we ended up shifting our architecture to Vanilla JS very close to the deadline and managed to get it up and running.

Accomplishments that we're proud of

Creating a POV Projection

This required 3 hours of marathon effort to create a shape that would conform to a normal point of view of a human. The bigger accomplishment was to getting the view matching the phone orientation. It required a lot of patience as we experimented with several methods, eventually one yielding great rewards.

What we learned

How to learn new things quickly is what we all learnt. We had not much experience with iOS or WebGL to plot a 3D map. But we adapted to the challenge and learnt how to work collaboratively and feed off each other. In ters of technology.

What's next for ThingFinder

We are trying to partner with organisations that have an infrastructure already in place. Cabs, Buildings, Old-Age homes and other places that require tracking of objects can benefit from our technology to find and locate things easily, saving time and money.

Built With

Share this project: