SmartROV is a Remotely Operated Vehicle (ROV) powered by AI. 

The goal of SmartROV is to conduct autonomous discovery sessions without the need of human control. The vision behind the project is to create a cheap ready to use rov that exploits AI to deal with all the challenges and limitations that drones have while operating underwater like:

• No GPS 
• Low visibility 
• Impossible transmission with radio frequency 

SmartROV is a small and low price ROV that can be used to run inspection and that continuously gather data related to the environment. It is using the following technologies:

• Computer Vision and CNN to detect fish species and to gather visual information related to the environment (i.e. kind of bottom)
• Sensors to gather information related to water quality, temperature, tides 
• SLAM (Simultaneous localization and mapping) to detect its position and to create a 3D map of the environment explored 
• DRL (Deep Reinforcement Learning) to be able to autonomously navigate and self levelling itself in the most efficient way 
• Edge Detection to improve visibility in low visibility environments

These features make SmartROV the first low cost AI powered rov ever. 

Why do we gather all these data? In order to make them accessible to everyone needs to run a research underwater topic. All SmartROV data will be gathered and stored on the cloud to make them accessible. Data will be of the following kind:

• Information about the water in the different locations (temperature, tides, direction of current, etc)
• Information about the bottom (kind, depth, composition, etc)
• Information about flora and fauna both qualitative (pictures) and quantitative (data extracted from all the pictures and videos) related to the kind of fauna and flora present, behaviour, information about the diferent location and the directions 
• 3D reconstruction of the visited sites in order to slowly create a high quality 3D map of the underwater ecosystem. 

This product will allow us to gather together data of different kinds and to make them publicy available to do a step forward in the research about the ocean that is struggling due to the extension of the site. 

But, in order to pursue a goal of this kind would be necessary to have thousands of SmartROV all over the planet, and this is impossibile from an economic point of view unless this product can be sold around the world to perform other tasks while gathering all the data. 

That is why we decided to start developing SmartROV as the first ROV to assist the diving centers all around the planet. SmartROV will be developed and sold as en economic platform for diving to:

• 3D build all the dive sites to show to the customers before the dive (this is currently done usually through a whiteboard)
• As a camera to take picture/video underwater to sell afterwards to the customers 
• As a tool for the captain of the boat to keep track on what is going on underwater and to prevent accidents 
• As an emergency tool to keep track of all the members of the group that are diving, sending an alert to the instructor/guide if a member gets lost or gets far from the group (through a computer vision algorithm that is currently under development). 

As a scuba diving instructor I know how much the previous points are really important for a diving shop, and I strongly believe that these specs make SmartROV a possible great product that can be widely sold. From a security point of view this could be a leap forward in this industry. 

And if we achieve this goal the higher is the amount of SmartROV around the world the more data they’ll gather and the more data researchers can use for their research. 

(From a privacy point of view a computer vision algorithm will make sure that all the data uploaded on the cloud are privacy complaints).

SmartROV could be the product that makes ocean mapping easier and achievable.

Link to your concept design and documentation

https://drive.google.com/drive/folders/1Znwgfep1aq4_LjF7ZScGinWiW9ffUgrh 

Link to source code of your solution or prototype above. (If you submitted a link to an online solution or prototype, or to a video of your solution of prototype, you must provide a link to the source code.)

https://github.com/LorenzoPinto04/AI-ROV