Alibaba Cloud and Longping High-Tech join hands in AI farming

Alibaba Cloud Computing Ltd announced a strategic cooperation with Longping High-Tech and CITIC Cloud to promote the application of the Alibaba ET to agricultural production, where ET would be used in breeding, seed selection, infrastructure datamation, farming management, selection of location and forecast of crop production outcome. This was announced at the Computing Conference held in Beijing last week.

Longping High-Tech is China’s seed industry leader, covering the development of hybrid rice, corn, vegetable and wheat seeds. Its research and development capability ranks among the top in the Chinese seed industry.

After the cooperation between Alibaba Cloud and Longping High-Tech, Alibaba’s ET service will become stronger in genome big data analysis, which will be able to perform the modelling of mass gene data-oriented characteristic dimension and discrete value. This approach will allow for an intensive exploration of the genome-soil-environment-phenotype relationship, so as to select simulated genome matching with desirable phenotype, which is a significant increase in the breeding efficiency. Due to limited size of the breeding bases and cost restraint in the past times, each year in a breeding project, only limited groups could be selected from thousands of available options for experiment, which was very inefficient. ET can help researchers to evaluate the breeding decision-making process and predict a hybrid species variety that will show optimum performance in the first-year experiment.

Besides breeding, ET will carry out datamation of planting bases, including acquisition of various structured and unstructured data in the course of agricultural production, which will be systemized and brought into the cloud service. Structured data include land area, planting species, time frame, soil sensor data, water and fertilizer data, as well as meteorological data. Unstructured data include UAV and camera photos.

ET will also play a part in prior selection of locations and species varieties. While taking into account the phenological, meteorological and altitude conditions, as well as the known origin of production of various crops, ET can predict best fit locations for specific crops, thus helping Longping High-Tech to select desirable locations. In one to two months before harvesting, based on the cross-crop and cross-season multidimensional data, ET can conduct an in-depth learning of any of the new crop varieties to predict yield and quality.