On January 18, the WIF2023 Innovation Pioneer Forum and Annual Meeting of Entrepreneur Members hosted by Taber was successfully held in Beijing. The theme of this forum is "Over the Hill", and more than 100 founders, managers and investors of technology companies gathered to study and judge the industrial economic trend and find new directions and new opportunities for the future development of enterprises under the new situation. As a leading enterprise in satellite application services, Shuhui was invited to attend the forum and participate in the industry roundtable dialogue.
At the opening ceremony, Liu Yuzhang, founder and chairman of Taber and president of Taber Research Institute, shared the "Top Ten Technology and Application Trends of Taber Think Tank in 2023", which focused on ten major fields such as generative artificial intelligence, humanoid robots, satellite Internet, data assetization, intelligent remote sensing, AI computing power, and low-altitude economy. For example, for generative AI, Taber Research Institute will generate a large number of subdivided model requirements with different parameters, different specifications, and different scenarios in 2024, and large models will usher in landing assessments. With the optimization of general large models, vertical large models of various industries have flocked to it. Only by mastering real-time industry data and training a large-scale model that is accurate, real-time, and deeply rooted in the needs of the industry can we truly gain a foothold. At the same time, the integration point of generative AI and decision-making AI is still in progress, and the landing solution needs to effectively combine generative AI and decision-making AI to jointly provide accurate and creative services for decision-makers, assist in optimizing business processes, and make intelligent decisions.
In the end of the book".Far up to the cold mountains-- Is the satellite big data market ushering in its second spring? At the industry roundtable dialogue, Lu Yuhang, deputy general manager of Shuhui Spacetime, Cheng Xiaoyang, general manager of Southern Remote Sensing, Bao Han, senior consultant of Tianhui Spatial Information Research Institute, Yao Liangjun, deputy general manager of Remote Sensing Division of SenseTime, and Zhang Xiaozhou, general manager of Aerospace Worldview, had a dialogue on the same stage, and conducted in-depth exchanges on "development prospects and challenges of satellite big data market", and focused on issues such as "satellite data assetization, remote sensing large model, and changes in the global remote sensing market pattern".
In the past two years, a number of remote sensing and meteorological models have emerged. In terms of remote sensing models, such as the Great Wall, Earth Boundary, and Tianquan models, and meteorological models, such as Pangu and Fengwu. These large models are exerting their advantages in depth and bringing more new possibilities to the industry. Lu Yuhang believes that the blessing of the large model can bring five possibilities to the industry:
High precision** and analysisThe meteorological model can improve the ability of short-term to medium- and long-term forecasting, especially in the prediction of extreme weather events. The remote sensing model can improve the accuracy and real-time performance of land cover classification, land use change monitoring, crop growth assessment, and disaster early warning.
Efficient data processingIn the face of massive satellite remote sensing image data, the large model can quickly realize data preprocessing, feature extraction, and intelligent interpretation, shorten the time cycle from data acquisition to information application, and improve data productivity.
Automation and intelligenceLarge AI models have strong self-learning and self-adaptation capabilities, and can adapt to specific tasks and application scenarios through transfer learning and fine-tuning without a large amount of labeling and training. In the field of remote sensing, AI large models can realize automatic and intelligent image recognition, target tracking, and scene classification. In the field of meteorology, AI large models can realize automated and intelligent weather forecasting, disaster warning, and climate change monitoring.
Cross-border integration and innovative applicationsAI models can break the boundaries between traditional fields and achieve cross-border integration and innovative applications. For example, the application of remote sensing technology and AI large models to urban planning and management can achieve intelligent urban planning and refined management. The application of meteorological models and AI technologies to the energy field can realize the optimization and scheduling of energy production and consumption.
Promote industrial upgrading and developmentLarge models such as remote sensing and meteorology can provide opportunities for digital transformation and upgrading of traditional industries. By applying large models to agriculture, forestry, environmental monitoring, urban planning and other fields, the digital transformation and upgrading of industries can be realized, and production efficiency and resource utilization can be improved. This advantage helps to promote the sustainable development of related industries.
Founded in 2003, Shuhui Spacetime is a state-level high-tech enterprise controlled by Shandong Land Development Group, which has been deeply engaged in the field of remote sensing AI for many years, and will release the multi-modal large model of natural resources "Great Wall" and remote sensing cloud brain DIEY in the first half of 2023.
The "Great Wall" model integrates data in various modalities such as space (i.e., vector), text, image (including image, **, etc.), and audio in the field of natural resources. Through the annotation learning of different modal data, it can effectively understand and generate natural resources investigation, approval, supervision and other services, and greatly improve the processing capabilities of element interpretation, change monitoring, and inversion.
Based on the "Great Wall" model, Shuhui Spacetime has developed a remote sensing cloud brain DIEY product. The product has made a number of technological innovation breakthroughs in the whole chain of satellite remote sensing applications such as satellite remote sensing element interpretation, image storage, image quality inspection, image repair, image preprocessing, image mosaic, intelligent interpretation, change detection, etc., and has formed more than 70 patents, and has been produced in more than 30 units in 17 provinces such as Guangdong, Xinjiang, Gansu, Shandong, Jiangsu, Ningxia, Heilongjiang, Henan, Shaanxi, etc., and is widely used in cultivated land "non-agricultural" and "non-grain" monitoring, high-standard farmland monitoring, forestry and grass monitoring, mine monitoring, Land supply monitoring, pine wood nematode research and judgment, moss monitoring, ecological GEP accounting and other fields.
With the gradual clarification and implementation of the data element policy and the prosperity of the market, Shuhui Space-Time has also had a clear positioning in the field of data circulation: it is not only a data product provider, a data product developer, a data governance service provider, but also a data asset quality evaluation service provider, providing high-quality spatio-temporal data product supply, a full range of spatio-temporal data governance services and quality evaluation services for the data element market, promoting the effective release of the value of spatio-temporal data and promoting the high-quality development of the digital economy. By the end of 2023, Shuhui Time and Space has cooperated with a number of data exchanges such as Beijing Digital Institute and Shanghai Data Exchange, obtained digital business qualifications, and has launched a series of data products.