The impact of artificial intelligence (AI) technology in advanced fields such as space exploration, unmanned vehicle manufacturing, or drug discovery is absolute. This is because human abilities are not enough to calculate complex formulas quickly and organize vast amounts of data in an instant.
However, the AI technology, which was thought to be applied only to such high-tech fields, is drawing attention as it is expanding its application range to primary industries such as forestry, fisheries, and livestock industries in recent years.
AI is expanding its application to primary industries
These days, the forest field is very emergency because of an epidemic called'Pine Tree AIDS'. This infectious disease, called'Pinewood nematode disease', quickly dries and kills pine trees, so once infected, the pine tree quickly becomes a dead tree.
Moreover, it is also highly contagious, so when a tree is infected, the pine forest in the area is often devastated. Therefore, to prevent infectious diseases, infected dead trees must be removed from the root and incinerated. The problem is that in order to find these dead trees, you have to go through them one by one.
The causes of dead trees vary considerably, such as climate change and soil type, even if it is not nematode disease. It is usually not difficult to read dead trees caused by various causes and trees that have been burned by falling leaves with aerial photographs.
However, with the introduction of the AI system to this difficult task, the reading results are changing drastically. It is the researchers of the National Park Service that introduced the AI system in the task of classifying dead trees.
These researchers used AI-based reading technology to investigate the area of Jirisan National Park. As a result of scanning an area of about 40 ㎢ over two days, it achieved the result of identifying about 55,000 dead trees.
Of course, in terms of accuracy, it is true that it is still inferior to the results confirmed with human eyes. It was found that it showed an accuracy of about 73% compared to the result checked by human eyes. However, when comparing the investigation period, the AI's reading results are overwhelmingly excellent.
Experts say that it takes about a year for people to search for a forest area of about 40 km2 and determine whether or not dead trees are dead. What would take a person a year or so was done by AI in just two days.
In the case of the Korea Forestry Promotion Agency, it has taken a step further and developed an AI algorithm that reads only dead trees that have died from re-nematode disease. If we can read only dead trees that died from re-nematode disease, it can contribute greatly to the prevention of infectious diseases.
The AI algorithm developed by researchers at the Korea Forestry Promotion Agency uses aerial photographs taken by drones that can identify objects up to 10 cm in size. The information on dead trees killed by re-nematode disease is built into a database from the photographs taken, and is used as training data for learning to improve the accuracy of reading.
It is also applied to livestock and fisheries to improve productivity.
In addition to the forestry industry, AI is having a profound influence on livestock and fisheries. The livestock industry is evolving to a level that predicts diseases in advance and streamlines livestock management by analyzing livestock activity and movement with AI through CCTV images installed in livestock farms.
The domestic livestock industry is much closer in quantity than the overseas livestock industry, but the productivity is quite low. The reason for this low productivity is due to lack of skills to manage livestock or poor ability to cope with epidemic epidemics.
The result of inadequate coping with infectious diseases led to a high mortality rate of livestock. In order to solve this problem, researchers at Korea Livestock Data, supported by the Ministry of Science and ICT, succeeded in developing an AI livestock management solution based on livestock behavioral data.
The AI livestock management solution is an optimized breeding system that can proactively respond to mass livestock management or livestock diseases, thereby contributing to the efficiency of the livestock industry by improving farm productivity by 30% and reducing mortality by 40%.
In particular, Korea Livestock Data, which developed the solution, aims to help advance the livestock industry in developing countries by exporting AI livestock management solutions to Southeast Asian countries where the livestock environment is poor. It also has a plan to show off Korea's livestock technology by exporting solutions to advanced countries such as Europe.
This solution is regarded as one of the achievements of the Digital New Deal that the public can experience. It is a representative example of digital transformation that innovates existing industries and creates new markets by fusing digital technology called AI with the livestock industry suffering from labor shortages.
The field where AI technology is actively applied as much as the livestock industry is the fishery industry. In the case of the National Institute of Fisheries Science, since 2016, it has been smartly managing fisheries by applying AI technology to the farm site.
Since the farm is equipped with various sensors, it is designed to automatically collect information related to the fishery environment, such as water temperature and dissolved oxygen, and transmit it to the farm operator in real time.
In particular, since it provides breaking news about red tide and breaking news related to various marine disasters, operators can manage the farm without staying at the farm. The smart fishery management system built by the National Fisheries Research & Development Institute is currently being tested at 10 farms in Tongyeong.