OpenOceanCam

The Problem

Studying behavioral change underwater today is expensive, difficult, impractical. Why? The current system is to use a Go-Pro with a battery pack which is set on 1 second timelapse to take images. This system's maximum capacity is limited to 65-70h of recording, including night time that has no usable information. Unfortunately, this limits the utilization of this technique in remote areas, where access is limited by the distance from the coast and the cost of repeated diving for deployment - which is close to 80 USD every 3 days including boat rental, manpower etc. Over a year this adds up to significantly increase the cost of monitoring the marine conservation process and reduces the amount of useful data collected by a large amount. The worst part - HOURS of manual post processing after the cameras have been pulled out of water. Recording the presence of and to photoID marine specimens is key to the conservation process as it helps monitor and quantify the success and failure of different conservation techniques/projects. Dr. Alessandro Ponzo: "A system of single or multiple cameras that could record directly to HDD or SSD with external power would allow conservationists to understand habitat use and movement of these endangered species at a much larger scale. "

Our Proposal

The Open Ocean Camera is an underwater camera that: 1) Runs 4 weeks underwater 2) Autosorts animals by genus as they are captured by the camera 3) Allows 200 combinations of shooting modes! The camera comes with an app which allows you to set the shooting schedule for the next three weeks. You can adjust the time-slots, ISO, shutter-speed, photo/video etc for the desired period. The camera also comes with an on-board graphical machine learning interface allowing you to easily train an ML network to identify and classify animals. Since this software is on the actual camera itself, it makes this calculation in real-time and separates the photo's/video segment where an animal is seen from the hundreds of empty frames captured across the 3 week recording period. This saves hours of post-processing. We are also working on: 1) A special rechargeable battery for the camera that can push battery life to OVER 1 month 2) A fixture on the outside that allows the camera to be pulled by boat for reef studies 3) A tether to the surface that allows wireless transmission of pictures back to shore!

We Assume that...

1) The camera's lens/size is suitable for a wide-range of marine monitoring needs

2) The easy machine learning interface will work effectively with a variety of marine species

Constraints to Overcome

1) The extension in battery life, memory of the camera. 2) The flexible photographing scheduling & camera parameter setting customization (ISO, Shutter Speed, Focal Length etc) 3) Easy, code free, drag-drop interface to allow deep learning of specimens 4) Tethering cable to water surface that allows global wireless control of camera

Current Work

Our product is almost ready after first round of iteration. We now want to set up an effective supply chain and network to to larger scale pilots with the product.

Current Needs

1) Funding (Looking for ~ 25,000 USD over the next 6 months to launch kickstarter and distribute cameras to conservation groups) 2) Manufacturing and Supply Chain - need to identify reliable sources of manufacturing.