The agricultural sector plays a pivotal role in the national economy and the achievement of Sustainable Development Goals (SDGs), particularly in attaining national food security. In the second quarter of 2023, the agricultural sector contributed approximately 13.4 percent of Indonesia's Gross Domestic Product (GDP) and employed around 40.2 million individuals, accounting for about 29.4 percent of the total workforce (BPS, June 2022).

Over the past five years, both the Global Food Security Index and Indonesia's Food Security Index have experienced stagnation with a declining trend, especially during the pandemic and the geopolitical conflicts involving Russia and Ukraine. Indonesia's Global Food Security Index (GFSI) falls below the global average, ranking 63rd out of 113 measured countries.

Food security encompasses the availability of quality food, adequate access to food resources for individuals and households, the availability of nutritious food items, and sustainable access to food. President Joko Widodo consistently emphasizes the importance of accurate and reliable agricultural data for policymaking, particularly within the agricultural sector and addressing national food security issues.

High-quality data, such as agricultural land area, productivity, and fertilizer requirements, are crucial for formulating effective policies that impact the populace. The recently conducted 2023 Agricultural Census is expected to provide a comprehensive understanding of agriculture throughout Indonesia, down to the smallest administrative level.

However, the current agricultural data collection primarily relies on administrative data and has not fully embraced objective measurement methods. To meet the need for high-quality, comprehensive, accurate, relevant, up-to-date, and detailed statistical data, a transformation in government statistical processes is required.

The use of various new data sources and the application of cutting-edge technologies such as Big Data and Artificial Intelligence (AI) are essential components of modernizing statistics with the aim of producing faster, more detailed, and accurate data and information.

Utilizing Big Data in Agricultural Statistics

Since 2018, the Central Bureau of Statistics (BPS) has employed the Area Sampling Framework (KSA), developed in collaboration with the Agency for the Assessment and Application of Technology (BPPT), to calculate the paddy crop area. This method involves merging data from paddy field maps derived from remote sensing technology, utilizing satellite imagery as a sampling framework, and Android devices for field observations related to crop growth.

The outcomes of this method are employed to predict potential crop areas in the upcoming months, which is crucial for anticipatory policymaking. Earth observation satellite image data can be easily and rapidly obtained, both through paid services like Radarsat 2 and freely available satellites like Landsat 8, Sentinel 1, and 2, which can observe hard-to-reach areas.

Other countries and international organizations, such as China, Australia, Colombia, the United States, and the Food and Agriculture Organization (FAO), have developed artificial intelligence to swiftly, accurately, and comprehensively estimate land coverage and use, agricultural land areas, plant density mapping, and crop type classification based on satellite image data.

The United Nations Statistics Division has formed The UN Global Working Group on Big Data, Satellite Imagery, and Geospatial Data Task Team to support the use of satellite imagery and geospatial data in generating official statistics for national policymaking.

In Indonesia, various institutions including the Ministry of Agriculture, BRIN, BPS, BIG, and universities have conducted research on utilizing satellite image data for land mapping, land classification, crop growth prediction, and agricultural production calculations.

Leveraging Artificial Intelligence for satellite image processing and automating agricultural statistics, when combined with official statistical data from surveys and censuses, can provide a comprehensive and integrated overview of Indonesia's agriculture. This can also rapidly monitor progress towards the Sustainable Development Goals (SDGs) down to the smallest level.

With fast, accurate, and detailed data, Early Warning Systems can be developed to monitor food availability, taking into account factors such as weather and pollution. The implementation of Artificial Intelligence with satellite imagery can also monitor forest land cover, deforestation, and periodic forest fires. The results of AI development through satellite imagery must be validated by examining on-ground conditions.

Moreover, it is crucial for statistical offices and government agencies to transparently and carefully implement AI, adhering to ethics, international standards, and the ten basic principles of official statistics. Challenges include access and costs associated with acquiring satellite image data, large data volumes, complex and varied data structures, and the need for robust computing infrastructure to process this data.

Collaboration among government, academia, businesses, the community, and media is essential within the national data ecosystem, particularly to support the strategic plan for AI implementation in Indonesia, including human resource development and infrastructure such as the National Data Center (PDN) and high-performance computing infrastructure.

To ensure data access and data sovereignty, Indonesia should consider having its own Earth monitoring satellite. Finally, synergy among policymakers in Indonesia is necessary to establish regulations related to Artificial Intelligence, not only in agriculture but across all sectors, ensuring safety, ethics, fairness, transparency, and accountability in its implementation.