Android Chicken Prediction Bot

Employer not named by the sourceRemote

Full Stack

Apply on the company’s site

Frontier is not the employer and does not collect applications.

About this role

PHP, Java, Mobile App Development, Android, Kotlin, Mobile Development, Android App Development, Data Augmentation · I want to turn the quirky idea of a “Chicken Road Prediction Bot” into a fully working Android app. The concept is simple: feed the app either live video or pre-recorded movement data of chickens near a roadway and have it predict if, when, and where a bird will step onto the road. The platform is strictly mobile and Android is the only operating system I need supported.

Here’s what I expect: • A lightweight Android application (Kotlin or Java) with a clean UI that shows a camera preview or data feed and overlays the real-time prediction. • An on-device model—TensorFlow Lite, PyTorch Mobile, or a comparable framework—that can run offline and outputs a confidence score plus an ETA for the crossing. • A simple stats panel logging every prediction with time-stamp so I can review accuracy. • Full source code, model files, and build instructions so I can re-train or tweak the logic later.

I’ll provide any chicken-movement footage I already have, but please plan on suggesting additional sample-collection or data-augmentation strategies if the model needs more variety. The project is done when the APK installs on a modern Android phone, processes a short test clip in real ti