Skip to content

Home

Satellite Data Utilization via Space Data on Demand
Space Data On-Demand is a technology that leverages AI and computing to generate customer-driven satellite data and to enable integrated analytics that fuse satellite data with terrestrial data, thereby solving challenges faced by industries and society.

The benefits of Space Data On-Demand

  1. Satellite Edge Computing
    • It becomes possible to quickly and accurately detect important insights from data acquired by satellites and transmit them to the ground. This enables new, near-real-time services that were previously unachievable.
  2. Observation Data Precision Enhancement Technology
    • AI improves spatial resolution, enabling high-precision predictions even in areas with limited observation data. For example, High-precision Precipitation Estimation Technology, which performs precipitation forecasting, can now achieve highly accurate precipitation forecasts that were not possible before.
  3. Large-scale Geospatial Data Processing Platform
    • By integrating satellite data with ground-based industrial data (such as port data, land transport data, factory IoT, POS data, and human flow data) using Satellite and Industrial Data Fusion Technology, we can visualize economic activity, societal dynamics, and global transport conditions everywhere on Earth in real-time.

Technology Overview

Target Industry/Users

  • Large enterprises with complex transportation networks, such as globally operating manufacturers, trading companies, and retailers, as well as the shipping companies, logistics companies, and warehousing companies that support them.
  • Service providers who acquire and analyze satellite data in near real-time and promptly notify results.

Challenges in Target Industry and Operations

  • Opacity of Global Transportation: Difficulty in accurately understanding and tracking their entire transportation network in real-time, especially for remote areas and maritime transport.
  • Delayed Response to Unforeseen Risks: When unexpected events such as natural disasters, port congestion, or geopolitical risks occur, it is challenging to quickly grasp their scope of impact and business implications, and to implement appropriate countermeasures.
  • Limitations in Prediction Accuracy: Conventional prediction models struggle to forecast the future of global transportation (e.g., estimated time of arrival) with sufficient accuracy due to complex intertwining factors, leading to lost opportunities and additional costs.
  • Limitations of Partial Optimization: While optimization proceeds for individual transportation modes or per-location, it is difficult to achieve overall optimization across the entire global transportation network, leaving inefficiencies and bottlenecks unresolved.

Technical Challenges

  • Lack of Wide-Area, High-Frequency Real-time Data Collection Capability: There is an insufficient means for continuous and high-frequency global data collection, especially in remote areas and at sea, that does not rely on ground infrastructure.
  • Lack of Heterogeneous Data Integration and Analysis Capability: There is a shortage of technology to seamlessly integrate diverse data formats and granularities, such as satellite data, weather data, and industrial data (AIS, port information, land transport data, factory IoT, POS data, human flow data), to derive meaningful insights.
  • Lack of Complex Spatiotemporal Causal Modeling Capability: There is a lack of AI technology that can accurately predict and simulate the dynamics of global transportation, where multiple factors are intertwined, and the impact of an event in one location on other locations.
  • Inflexible Space Platform: There is a lack of technology to realize a high-performance, high-reliability environment that can be easily programmed according to user needs.

The benefits of Space Data On-Demand (Detailed version)

  • Proactive Management through "Earth Observation":
  • Why it leads to value: Large-scale Geospatial Data Processing Platform continuously processes high-frequency satellite observations and real-time ground data with Satellite and Industrial Data Fusion Technology, capturing changes in the business environment on a daily and hourly basis. This allows for immediate and comprehensive understanding of global transportation bottlenecks, disaster damage, competitor production trends, and more.
  • Value: It transforms the quality of decision-making itself from "post-event response" to "proactive management" that anticipates signs and takes preemptive action, directly leading to the avoidance of lost opportunities and cost reduction.
  • Minimizing Business Impact with Overwhelming Prediction Accuracy:
  • Why it leads to value: In addition to high-precision disaster prediction by High-precision Precipitation Estimation Technology, Satellite and Industrial Data Fusion Technology integrates rich semantic information extracted from satellite data and diverse industrial data into a rich spatiotemporal graph, which is then learned by STGNNs, providing high-precision future predictions that were impossible with conventional statistical models.
  • Value: It achieves the avoidance of container demurrage and detention fees, which can amount to tens of thousands of dollars per day, the prevention of serious secondary damages such as production line shutdowns and lost sales opportunities, and the maximization of cost efficiency through optimal inventory placement and transportation planning.
  • Sustainable Competitive Advantage through "Global Optimization" of Entire Global Transportation:
  • Why it leads to value: Large-scale Geospatial Data Processing Platform models the entire global transportation not as fragmented individual elements, but as a single integrated "geospatial graph" on a unified grid. By using STGNNs on this graph, it simulates the impact (ripple effect) of intervention in one part on the entire system, deriving a "globally optimal" solution that maximizes system-wide performance, which can never be reached by the sum of individual partial optimizations.
  • Value: It establishes a sustainable competitive advantage by simulating the movements of all stakeholders such as ports, land transport, and warehouses, and proposing operations that maximize the efficiency of the entire ecosystem.
  • Satellite Monitoring:
  • Why it leads to value: For applications that could not be processed until they were released on the ground due to advanced data processing, immediate processing on the satellite allows important results to be notified quickly.
  • Value: Enables applications that require immediate notification, where meaning would be lost after a few hours.

Fujitsu's Technological Advantage

  • Depth of Data Fusion with Satellite and Industrial Data Fusion Technology: Satellite and Industrial Data Fusion Technology, which not only superimposes satellite and industrial data but also deeply learns and infers their temporal and spatial causal relationships using geospatial graph structures and STGNNs (Spatiotemporal Graph Neural Networks), is a unique strength unmatched by other companies. This enables more accurate and reliable prediction and optimization.
  • High-Precision Prediction in Areas with Sparse Ground Observation Networks with High-precision Precipitation Estimation Technology: High-precision Precipitation Estimation Technology, which enables high-precision precipitation forecasting using only satellite data, has a superior advantage over other companies, especially in predicting natural disaster risks in areas with insufficient ground observation data, such as developing countries and remote regions.
  • End-to-End Integrated Architecture: Large-scale Geospatial Data Processing Platform's integrated architecture, which provides an end-to-end solution from data collection to preprocessing, feature extraction (utilizing foundation models), modeling (STGNNs), and concrete problem solving, unlike competitors' "point solutions," enables true "global optimization."
  • Safety in High-Load Processing: By optimizing processing efficiency and reducing power consumption, it detects and corrects errors due to cosmic rays, achieving high reliability.

Use Cases

  • End users:
  • Scene 1: Risk Assessment during Global Transportation Planning and Execution
    • How to utilize: When planning coffee bean transportation, display a map of the coffee farm area in Brazil on the Large-scale Geospatial Data Processing Platform screen. By using High-precision Precipitation Estimation Technology's precipitation forecast, proactively identify areas with high flood risk. By selecting alternative routes suggested by Large-scale Geospatial Data Processing Platform, delay risks due to natural disasters can be avoided even before transportation begins.
  • Scene 2: Responding to Unexpected Situations during Maritime Transport
    • How to utilize: As a vessel in transit approaches its destination, Tokyo Port, Large-scale Geospatial Data Processing Platform analyzes satellite data (optical, AIS) with Satellite and Industrial Data Fusion Technology and detects severe port congestion beyond official information. By quickly deciding on the optimal option from multiple alternatives (e.g., rerouting to a port in Kansai) suggested by Large-scale Geospatial Data Processing Platform, considering cost, delay, and business impact, critical business losses can be avoided.

  • Scene 3: Strengthening Global Transport Resilience and ESG Management

    • How to utilize: Large-scale Geospatial Data Processing Platform continuously monitors supplier factory operating status, environmental risks along transport routes (e.g., methane emissions, deforestation), and port congestion using Satellite and Industrial Data Fusion Technology. This increases the transparency of global transportation, provides verifiable evidence for ESG reporting, and contributes to strengthening global transport resilience and sustainability by early identification of potential risks and proactive measures.

  • App Developers:

  • Scene 1: Feature Expansion for Existing Global Transportation Management Systems
    • How to utilize: By leveraging APIs provided by Large-scale Geospatial Data Processing Platform (Feature Extraction API, STGNN Forecast API, Geo-enrichment API, etc.), real-time risk prediction based on satellite data, high-precision ETA prediction, and alternative route suggestions can be easily integrated into existing SCM/ERP systems. This enhances the system's value without the need for in-house satellite data processing or advanced AI model development. -Scene 2: Development of Geospatial Intelligence Applications for Specific Industries
    • How to utilize: Using Large-scale Geospatial Data Processing Platform's Satellite and Industrial Data Fusion Technology as a foundation, new geospatial intelligence applications tailored to specific industry needs can be rapidly developed. For example, in agriculture, monitoring crop growth and predicting yields; in finance, geospatial risk assessment for investment targets. Since Large-scale Geospatial Data Processing Platform centrally provides common functions like data integration, AI inference, and API provision, developers can focus on application-specific logic development.
  • Scene 3: Integration with Digital Twin Construction and Simulation Platforms
    • How to utilize: By linking the geospatial graph data and prediction results generated by Large-scale Geospatial Data Processing Platform with digital twin platforms such as Fujitsu's Social Digital Twin, applications capable of advanced use cases such as more detailed simulations, policy-making support, and urban planning optimization can be developed.

Case studies

  • (Future) Global Transportation Optimization PoC:
  • Assuming the transportation of high-quality coffee beans from Brazil to Japan, we will demonstrate flood risk prediction at the production area by High-precision Precipitation Estimation Technology, detection of Tokyo Port congestion from satellite data by Satellite and Industrial Data Fusion Technology, and avoidance of business losses by rerouting to an alternative port. We plan to demonstrate the potential for avoiding losses on the scale of tens of millions of yen, proving Large-scale Geospatial Data Processing Platform's value.
    • Developing a solution that analyzes container yard congestion at ports in real-time using satellite SAR imagery and AIS data via Satellite and Industrial Data Fusion Technology. AI optimizes container loading/unloading schedules, contributing to reduced truck waiting times and improved berth utilization.