Final round winners have been announced. View Results

Team

orcast

Project Concept

ORCAST: Immersive Orca Encounter Platform

Project Overview

ORCAST combines marine science research with real-time environmental data to support orca conservation and improve whale watching experiences. The platform integrates behavioral forecasting, interactive mapping, and AR-guided field viewing to help users plan encounters while contributing sighting data to research.

What’s Already Built and Working

The scientific infrastructure is complete. ORCAST operates as a production system at orcast.org with comprehensive data and ML infrastructure. The backend integrates neuro ethology lab research analyzing kinematic and depth time series from DTAG biologging data with real-time environmental monitoring.

BigQuery data warehouses store DTAG deployments, expert marine biologist annotations, and environmental correlations. Cloud Run services provide real-time behavioral prediction and feeding strategy classification with sub-500ms response times. Firebase manages live sighting reports and environmental data fusion from NOAA sources, historical sightings, and vessel traffic monitoring.

The system classifies feeding strategies from kinematic patterns and integrates environmental feedback from experimental conditions. This enables behavioral forecasting for feeding, traveling, and socializing behaviors while tracking conservation impact.

The Hackathon Challenge

With robust backend infrastructure operational, the hackathon focuses on creating user-facing experiences that make this scientific platform accessible. The goal is building a working demo by 6pm Day 1.

Day 1 involves creating an Angular application with Material Design connecting to existing orcast.org APIs. Core mapping with Google Maps for probability visualization and basic research analytics with Chart.js. Simple device orientation for AR field-of-view concept demonstration.

Day 2 is demo preparation and final polish only.

Three Core Experiences to Build

Research Analytics Dashboard is an Angular web application where researchers visualize feeding strategy analytics, explore environmental correlations, and track sighting success metrics across locations and conditions.

Interactive Map provides a web and mobile interface displaying live probability heatmaps across San Juan Islands, historical patterns for seasonal/tidal trends, and viewing locations with success rates. Social sighting reports with photos create a community knowledge base.

AR Field Viewer uses Progressive Web App technology and device sensors for compass-guided viewing. Provides real-time probability overlays based on viewing direction, distance/bearing indicators for approaching pods, and behavioral context. Encounter documentation with automatic location tagging contributes data back to research.

Technical Implementation

Development stack uses Angular with Material Design, Google Maps for mapping with real-time overlays, Chart.js for analytics visualization. Device integration through compass, GPS, and camera APIs enables AR functionality. Direct API connections to orcast.org provide backend services.

With BigQuery ML pipelines, Cloud Run services, Google Cloud Vertex AI, Firebase hosting, DTAG data integration, environmental data fusion, and Redis caching operational, the team focuses entirely on user experience rather than backend complexity.

Team Contributions Needed

Frontend developers handle Angular component development, Progressive Web App implementation, and field condition optimization. Data visualization specialists design research analytics dashboards and interactive behavioral pattern charts. Mobile/AR developers integrate device sensors, field guidance overlays, and outdoor performance optimization.

UX/UI designers map user journeys and create research and consumer interface design. Marine science consultants validate behavioral classifications and develop conservation messaging.

Impact and Applications

The Southern Resident Killer Whale population has declined to 73 individuals. Each encounter provides valuable data contributing to protection efforts. ORCAST enables whale watching trips to contribute to conservation while improving encounter success through scientific prediction, crowdsourced data collection, and real-time conservation education.

This approach helps ensure continued opportunities to observe orcas in their natural habitat while contributing to the science that supports their protection.

Entry

Status: In Progress

Last saved: July 20 at 11:18 AM EDT

View Entry

Team Roster

Message board not available for this team yet.

Gil J Raitses Team Lead RSVP Approved

Founder at Fifth Avenue Studios

I’m an artist, filmmaker, culinary poet, dog papa and drag student oops I meant grad student. I work with a neuroethology lab in Syracuse where I also did some of my undergrad studies in Film… some 20 years ago. I’ve gotten into data modeling and visualization from the arts/hippie door, and it’s been a great trajectory for my creative research. And it’s opened the door for me as well into basic research, which I’m finding many intersections between. So I guess I’m at a very happy crossroads that just keeps getting more interesting.

I’m interested in immersive analysis tools for exploring pattern relationships using spatially grounded, perceptually-driven methods. I focus on cross- and intermodal signal processing, exploratory visualization, and systems that link sensory dynamics with real-world data. I’m interested in building workflows that reveal emergent patterns, support nonlinear inquiry, and let users move fluidly between abstraction and physical context.

Mechanosensation reinforcement learning analysis with drosophila larvae. I develop analysis classes for the lab. I’m working personally on a narrative TTS add-on module to interpret experiment data.

Dante Mienie RSVP Approved

Software Developer at Recent Grad

Hi! I’m Danté Mienie, a recent Computer Science graduate from the University of Arizona, where I also minored in Statistics and Data Science. I love trying new things and learning more. I am especially curious and excited aboutthe intersection of tech, finance and AI. In my senior year, I worked with aresearch lab that created a medical case simulation web app that integrated AI to help medical students with different aspects such as suture analysis and patient-doctor practice.

Finance, quant, web development, AI, app development

I’m working on an FastAPI investor ai that gives suggestions on a couple of tickers to consider investing in and why based on the marked. I’m also building an IOS music social media app that I hope to publish to apple store soon.

AI Tinkerers - New York City Hackathon