Fishackathon 2016 problem statements 7 & 9
Currently, there is no integrated tool to visualize and process all the information available about vessels, marine protected areas (MPA) and illegal blacklists. In particular, it is difficult to detect illegal fishing activities in marine protected areas since it is costly for coast guards to constantly watch out for boats and to send out helicopters over the seas.
This project aims to use the power of the crowd to quickly identify illegal boats in MPAs as well as providing a ‘one-stop-shop’ dashboard app to process data from all different sources.
How did we use satellite data, statistical analysis and gamification to tackle this problem?
The two-part solution
- A simple, addictive mobile game – tap as many boats as possible to win points!
- A Dashboard that empowers the coastal law enforcement and the public to digest all the information available. Includes heat maps, data tables of information and legal information all in one place.
Satellite images of MPA's are cut up and set as the background for the game. The user has to tap as many boats in the frame as quickly as possible, in a high-paced, addictive setting! If the system detects that the tap is OUTSIDE the randomized color boat (see video), it recognizes it as a potential illegal boat in the satellite image.
All the tapped data gets processed through calibration and statistical models to determine whether if the tap is indeed an illegal boat. Better users’ results get weighted higher to improve and filter the data.
Each background image covers an area of 500m2 – imagine how much area of the sea can be covered in each 10 second game!
The game mechanics can be improved by adding a musical element, improving the calibration processes for the images and linking it to a game engine for better performance.
The demo dashboard currently has a heat map which visualizes the data collected from the Tap-a-Boat game. It shows the coast guard where the potential illegal boats are and where to investigate. It has information about other MPAs, plugs in GPS AIS data from nearby ships in that area to improve accuracy and a social communication system.
All these tools are modular and other data sources from any other organization or data set can be easily plugged in into an extra widget!
It can be adapted for different countries, languages and needs for each community. It can potentially also be opened to the local public community so that it empowers them to take action as well. There are problems linked to corruption which prevents action from being taken. If the public are aware and informed about the illegal fishing problems, they can help tackle the issue.
From a technical point of view one can distinct different elements. The current mobile game is running as a web app in the browser. It uses the canvas HTML element to fill the game. Furthermore we expect to be able to develop a hybrid-app quickly.
We can use mongoDB and node.js to implement the back-end side of the app.
- HTML 5 (canvas elements)
- Sentinel 2 Satellite images
- Integrate a machine vision algorithm to aid the image recognition.
- Improve the game mechanics to make it more addictive and simple to use
- Speak to experts and scientists in the field to better understand the problem and therefore the solution
- Create native apps (as opposed to this webapp) to improve accessibility and performance
Try it out! (on Chrome and your mobile)
Game prototype – http://fish.eu-gb.mybluemix.net/game.html
Dashboard Prototype - http://fish.eu-gb.mybluemix.net/
Git repo - https://github.com/eunjinlee8/tapaboat.git