Ganos is a spatiotemporal big data engine system launched by Alibaba Cloud. This engine is used to manage spatial geometric data and spatiotemporal trajectories.

Ganos is compatible with open source ecosystems such as GeoMesa and GeoServer. Ganos provides efficient built-in algorithms, such as algorithms for spatiotemporal indexing and geometric algorithms for calculating topological spaces. You can use Ganos together with LindormTable of Lindorm and the serverless Spark engine in Data Lake Analytics (DLA). This way, you can perform distributed storage and data analysis. Ganos is suitable for multiple scenarios. You can use Ganos to store, query, analyze, and mine spatial or spatiotemporal data.

  • You can use Ganos to process spatiotemporal geometry data. For example, you can use Ganos to display and model spatial geometry data and spatiotemporal trajectory data.
  • You can use Ganos to manage spatiotemporal geometry data. For example, you can use Ganos to create, insert, index, query, or delete data.
  • Ganos can be used together with the Spark engine in DLA to perform data analysis. Ganos is compatible with GeoSQL that is defined by Open Geospatial Consortium (OGC).

Ganos can be used at the data layer of DLA to process remote sensing images. For more information, see Spatiotemporal raster.

Scenarios

  • IoT in transportation

    Ganos allows you to write trajectory data in high concurrency mode, store historical trajectory data, and specify a time range to query spatiotemporal data. For upper-layer scenarios, you can use Ganos to schedule transport capacities, organize carpooling, forecast supply and demand, and analyze heatmaps.

  • Big data processing for vessel tracking

    Ganos allows you to write automatic identification system (AIS) data in real time, store historical vessel trajectory data, and playback trajectories in specified spatiotemporal ranges.

  • IoT

    IoT devices generate time series data and spatial data. For example, a large number of connected vehicles generate trajectory data that contains time and space information in Internet of Vehicles (IoV) scenarios. In this scenario, you can use Ganos to monitor vehicle data in real time. For example, you can track vehicle trajectories and identify whether vehicles deviate or enter restricted areas. Ganos also allows you to query spatial data in real time. You can query the trajectory of a vehicle during a period of time. You can also query the vehicles that enter an area during a specific period of time. Ganos is used together with frameworks such as Spark to perform big data analysis. This way, you can perform spatiotemporal queries and analyze heatmaps.

  • Intelligent logistics and takeaway delivery

    In logistics and takeaway delivery scenarios, users need to track delivery vehicles and riders. You can use Ganos to estimate the delivery time. In this scenario, a cloud system for data analysis must support high concurrency for fast write operations. The system must also plan vehicle routes in real time and monitor vehicle deviations.

Benefits

  • Lindorm uses a cloud-native NoSQL database architecture and provides a service that is ready for immediate usage for processing big data.
  • You can use Lindorm to store petabytes of data. Lindorm supports highly concurrent write operations and handles tens of billions of data entries in a few seconds.
  • Lindorm supports separate storage of cold data and hot data and uses efficient compression algorithms. You can use Lindorm to store large volumes of data at a low cost.
  • Lindorm is used together with the Spark engine in DLA. This helps you build large spatial data repositories and platforms that are used to analyze spatial big data.
  • Ganos is compatible with Open Geospatial Consortium (OGC) standards.

Scenarios and architecture

  • Big data platform for vessel tracking
    • Scenario
      The AIS collects hundreds of millions of trajectory data records each day. The system tracks vessels in real time. Ganos allows you to perform spatiotemporal queries and time series queries on trajectory data in a few seconds. Ganos provides the following features:
      • Use geofencing to determine whether you are entering or exiting a geographic region in real time.
      • Specify a region and a time range to replay the trajectories of vessels.
      • Specify a time range to replay the trajectories of vessels.
    • Architecture
      • Build a data warehouse to store AIS data that is cleansed and processed on Spark.
      • Ganos stores the data that is cleansed. After the data is stored, you can query the data from a frontend application. For example, you can query geo-fencing data or replay the trajectories of vessels.
      • In this architecture, data queries can be performed in a few milliseconds and data cleansing can be performed in a few minutes.
  • Location tracking
    • Scenario

      Collect the trajectory data of vehicles in real time. You can collect the trajectory data of more than 400,000 vehicles and write up to 30,000 trajectory points per second. You can navigate vehicles in real time, perform geo-fence queries, and calculate the similarity between trajectories.

    • Architecture
      • Use Ganos to store all historical trajectory data.
      • If the data volume of historical trajectories is large, the system automatically dumps cold data to Object Storage Service (OSS). This reduces costs by more than 70 percent.
      • Query tens of billions of trajectory data records within sub-seconds. You can query trajectories by time range or by space.
      • Integrate into Spark to provide the analysis feature for trajectory data.
  • Manage remote sensing image data and enable an intelligent service
    • Scenario

      Store, query, and analyze remote sensing image data and enable AI for remote sensing.

    • Pain points
      • The data volume of remote sensing images is large and grows at a fast rate. The scaling process for storage resources is difficult to manage. This increases costs and makes management complex.
      • If you want to quickly display data, divide static data into segments. This traditional method discards metadata and cannot meet the requirements for AI-based data analysis.
      • Integrated queries and analysis from multiple spatial data sources are difficult.
    • Architecture
      • Use OSS, Apsara File Storage NAS (NAS), and storage area network (SANs) to store raw data and separate cold data and hot data.
      • Use PolarDB to store image metadata and use Lindorm to store tile data. This helps ensure that storage resources can be scaled.
      • Use a cloud native architecture and break up data into blocks for storage and management. Ganos retains the pixel data of images and is suitable for data analysis and computing.
      • Ganos is used together with Spark. This helps accelerate extract, transform, load (ETL). Ganos also helps you analyze and process spatial data from multiple sources.