Cross-domain Advance Technology Support Teams Launched to Meet the Trials on Agriculture

Hung, Chia Hung（China Productivity Center Smart Agriculture Promotion Department ）

Faced with increasing threats like the extreme weather, aging workforce, and pests and diseases, agriculture practitioners all seek to meet the challenges by incorporating the use of sensing technologies, intelligent machine devices, human-machine collaborative aids, the Internet of Things, massive data analysis, AI, and other cross-domain advance technologies to achieve the ideal of smart agriculture, and it has become the future development trend of global agriculture. Taiwan's agriculture cannot stay out of this trend. The sustainability of agriculture with environmental protection and sustainable production is a serious lesson for Taiwan's agriculture.

In the past ten years, digital development is sought after by all industries to make the most efficient use of limited resources. With the incorporation of information and communication technologies and sensing technologies, the agriculture sector needs to analyze the variability of different environments and formulate corresponding solutions based on the diverse data collected. This is supposed to administer appropriate resource input and treatment, avoid waste of resources and excessive pollution, and strike a balance between the pursuit of economic benefits and ecological environmental protection through the use of lean agricultural management.

Among all the cross-domain advance technologies, artificial intelligence (AI) makes use of deep learning algorithms and automatic image recognition technologies to provide solutions to the threats faced by global agriculture through early warning on climate impact, precision fertilization, irrigation and weeding and other planting management, automatic identification of crop maturity to assist in harvesting, identifying diseases and insect pests, developing new varieties by gene sequencing, etc. This even develops a series of innovative services and achieve the comprehensive upgrading and transformation of agricultural resources and agricultural manpower.

Another example is the use of big data analysis, which has also brought opportunities for changes in agriculture. In the era of information explosion, the agricultural industry cannot ignore the trend. The use of big data in agriculture involves farming, sowing, fertilizing, and insecticide, harvesting, storage, breeding, etc. and constitutes analysis and exploration in a cross-industry, cross-profession, and cross-domain manner.

To sum up, in order to establish smart agriculture for subsequent practical applications, several core technologies need to be integrated, including: data collection and transmission on Internet of Things, AI big data analysis, cloud automation control, etc. Below is a simple introduction to these usages.

1. Application of Internet of Things in Agriculture

The remote monitoring capability of IoT is widely used in industrial machinery maintenance. The function of IoT devices to collect detailed and real-time diagnostic data, however, makes it a powerful means in the field of agriculture.

Taking indoor growers and nurseries for instance, IoT devices can help monitor and manage plant production. In this scenario, on-line devices continuously monitor environmental conditions, such as temperature, humidity, sunlight, and soil, with the data collected, processed, and organized to help growers analyze long-term environmental trends. In addition, IoT devices also can also make automatic responses to environmental data. For instance, low soil moisture detected automatically triggers the watering function of the irrigation systems. This improves productivity and reduces labor costs. As with other IoT applications, the end result is a more efficient, more productive and lower cost system.

More than that, the IoT virtually reversed the existing business model. Take the agricultural machinery industry in the US for instance. In the past, enterprises simply sold their products. For example, agricultural machinery companies sold agricultural machinery. With the step-by-step incorporation of IT technologies, such as computers, communication equipment, and back-end control systems, the product strategies of the enterprise begin to change, from independent products to a system composed of closely related products, and finally to a system composed of a series of product systems. Now, the company's products are redefined again and again, and the industrial boundaries will continue to expand. By presenting one or more of these value points, IoT technology proves to be of great value in various applications and domains. With its development, there will undoubtedly be more and more novel and innovative applications in many industrial applications.

2. Application of big data analysis in agriculture

Digital agriculture includes three main elements: network and machine communication, cloud computing, and big data. Cloud-based data storage through the Internet of Things (IoT) enables new analysis methods to obtain new information from existing data. This is when big data give people new advantages. In terms of the agricultural market demand, big data can be used to help plan agricultural production, predict market demand, and assist in agricultural policy making, so as to achieve goals including risk avoidance, production and revenue increase, and transparent management.

For instance, the Council of Agriculture has the collaboration project with the Fiscal Information Agency of the Ministry of Finance on the analysis of agricultural product consumer behaviors by using data from electronic invoices and data science. These include the data from electronic invoices, whole sale prices, retail prices in city markets, prices in traditional markets, weather and public opinions, etc. Big data analysis tools are developed by using the data science to understand the consumption market trends of 15 agricultural products, such as cabbage and bananas and to establish retail price estimation models. Historical data are used to establish forecasting models to further simulate the impact on vegetable prices by factors like different paths, intensities, and rainfalls in typhoons, so as to be in advance control of price fluctuations. In addition, big data applied in the area of agricultural production would mean the use of sensors to collect climate and soil big data to provide farmers with optimal planting decisions, thereby assisting in effective management of the farmland and reducing agricultural costs.

3. Cloud-based automation control

Cloud-based automation control mainly transmits data to the cloud/local database by integrating various wired/wireless sensing devices and automation control systems, allowing the operators to perform remote real-time monitoring. Take smart greenhouses for instance, information and communication technologies are used to monitor multiple farmland data, such as water volume, fertilization, soil temperature and moisture, etc. Remote control technologies are then used to reduce the impact of external weather conditions. For instance, the greenhouse automatic adjustment function helps achieve favorable growth conditions for crops, accelerate the production cycle, ensure the stability of the output, enhance the flexible adjustment of the production seasons and improve the efficiency of greenhouse cultivation.

Other countries, such as the United Kingdom, use big data to integrate precision agriculture to achieve full broadband coverage in rural areas, platform and access channel setting, and integration and analysis of agricultural production information. In Germany, digital agriculture is developed for agricultural solutions with the use of various technologies of sensors and big data analysis, in which agricultural producers can monitor on their personal computers the intensity of sunlight, soil moisture content, and fertilizer distribution in real time. In summary, cloud-based automation control is the current world trend of smart agriculture, but the collection of a large amount of structured and unstructured data for various agricultural subjects will be a key challenge for all countries to develop smart agricultural solutions in the future.

Regarding the difficulties and potential challenges facing Taiwan's agriculture, the agriculture needs the support of science and technology, and science and technology should serve agriculture. At present, agriculture and technology are far apart from each other, and integration is in high demand to make agricultural development future-oriented. Information and communication technology (ICT) plays just a supporting role, and our goal is to make agriculture, our protagonist, become more intelligent with added value. Agriculture is different from the manufacturing industry. In manufacturing, all the processes are done in factories, but this is not the case in agriculture. Smart agriculture can complete 80% of work in production, but farmers still have to complete the rest 20% of the job. In the era of explosive growth of data information and ICT, the application of cross-domain advance technologies in agriculture requires more careful thinking of Taiwan government officials and agricultural practitioners in seeking breakthroughs.