Photo credit: Marine Biologist Research, HKU
Global marine and coral crisis
Nowadays, climate change has become a critical crisis that humans have to address. One Sustainable Development Goals (SDGs) of the United Nations is to combat climate change. Greenhouse gas emission causes sea temperature to rise and hence acidify the ocean, putting corals around the world at risk.
Coral bleaching is occurring on a global scale
A clear example is the Great Barrier Reef in eastern Australia, the corals there have been suffering from bleaching at an accelerating speed. Eventually, corals will die en masse and directly hit the marine ecosystem. With fewer corals to be kept at the bottom of the food chain, marine lives will be deprived of food and cannot survive . As the food supply from the ocean depletes, humans eventually suffer as well. If we don't act now and protect corals, the ocean and human beings will share the same threat, and eventually the same fate.
The coral ecosystem in Hong Kong
Beautiful coral at Basalt Island, Hong Kong. Source: Hong Kong Economic Times
As part of the global ecosystem, Hong Kong, the small island we live on, is not immune to the above hazard. According to statistics, there are over 80 species of corals living in Hong Kong at great diversity. If you have ever dived at Hoi Ha Wan or Sai Kung, you will be fascinated by the colorful corals there. However, there is also a growing concern over the problem of coral bleaching in Hong Kong, and it is particularly serious in the waters around Sai Kung. However, to save them, we first need to know where they are by stock taking of their species and numbers. This is never easy with conventional technology by scanning a large area of the sea floor.
CoralBot was born!
In light of this, MakerBay Foundation developed an automated marine robot, CoralBot, to solve the above problem (Click here for the project details).
CoralBot is equipped with 5 video cameras on an autonomous boat that can scan the ocean floor at a rate of taking 36,000 thousand photos per hour, providing millions of photos of the underwater world and its corals.
CoralBot developed by MakerBay Foundation
AI image recognition to taking stock of coral
Blueinno was the technical consultant for this project and worked with MakerBay to develop an artificial intelligence (AI) model by tagging coral photos, followed by performing object recognition with machine learning. Such method allows the computer to distinguish different coral species by itself, and translated the results into a systematic database consisting of different coral species. The database will serve as the foundation for providing the data and analysis to further facilitate the development and implementation of coral conservation projects.
Object recognition model building process
With the completion of the AI object recognition model, Blueinno also provided training to the MakerBay Foundation colleagues and researchers to ensure that they can operate the system, so as to execute the conservation project effectively and efficiently.
Educating the public about AI and conservation: Let’s do it together!
Of course, we also understand that without public support and participation, the conservation project will not be able to move forward. Therefore, it is important to raise public awareness of coral conservation and to gain their support. MakerBay Foundation and Blueinno hence took a step further by collaborating with the Ocean Park to organize a series of coral image recognition courses for 100 teachers and students from 30 secondary schools in Hong Kong from November 2020 to April 2021.
Through Blueinno's tailor-made AI course, participants learnt how to use Google Teachable Machine and saw the demonstration of YOLO in AI object recognition. We also invited them to build a big coral database together. The course was well received. Not only did the participants learn a lot more about AI knowledge and technology, but also were more aware of marine conservation and their roles in the global ecosystem.
The Blueinno team explained to the school teachers and students how AI performed object recognition.
Conclusion
In conclusion, AI applications and robotic engineering help us to cope with many social and environmental problems human beings are facing nowadays. A case in point is coral conservation in response to the marine crisis. Through object recognition, AI helps us quickly recognize and count the species, number, and location of corals at risk in Hong Kong, and create a big database to save them. The same technology can also be applied to other forms of conservation or applications such as counting medical instruments. Therefore, our new generation needs to learn and master AI technology early to innovate for a better future for human beings.
Blueinno has been committed to promoting technological knowledge by teaching programming skills and its applications, in an endeavor to turn our new generations into social innovators and entrepreneurs who can contribute to the human race. Blueinno also teaches a variety of courses covering cutting-edge technologies and knowledge such as machine learning, object recognition, blockchain and cryptocurrency, through a series of hands-on projects that allow students to practice what they learn.
Interested students and parents can visit the following pages or contact our teaching team.