Transforming the complexity of the real world into a virtual world (sponsored)

This story is brought to you by Presagis.

An explosion of data from a variety of connected devices and possibilities will see our world generate 10,000 times more data than already exists. Agencies, companies and countries are constantly collecting more data than they can process and analyze, impacting the need to make critical and complex decisions.

Whether for the defense, geoINT, infrastructure management or utility sectors, these growing volumes of geospatial data are driving new requirements for data analysis, 4D visualization, simulation and urban terrain modeling.

VELOCITY 5D — or V5D — can examine all information, make it clean, organized, and normalize data to create competitive advantage.

V5D is a software ecosystem that fundamentally changes the way users manage the production of digital twins and large virtual environments. V5D is able to exploit Big Data using:

• GIS data extraction, AI processing and labeling, multi-source fusion and inference to create simulation-ready instrumented data

• Continuous updates as new data becomes available

• Timeline – from historical data to present and future data (inferred or hypothetical scenarios)

• Data distribution from local server to cloud offering • Flexible and customizable architecture to meet specific customer requirements

V5D was recently used in two case studies that demonstrate the power of the digital twin creation pipeline.

Case Study A: Global Database

In 2020, Presagis participated in a project with a major US government agency to create a simulation-ready database (synthetic environment) of the entire planet with a particular focus on the continents of Asia, Europe and Australia. North America. To do this, Presagis used V5D. For training purposes, the customer needed both a global vision of operations in flight at high altitude, and high resolution areasaround the world when approaching an airport.

The project requirements were to collect high resolution satellite imagery, footprints of buildings, roads and vegetation to generate a high quality simulation database. Depending on the zone type, the database specifications varied. For example, global areas required medium resolution satellite imagery with some additional features, while specific areas of interest required high resolution imagery with geo-specific content models to help pilots familiarize themselves with their surroundings. especially airports.

A manual process for this project would involve years work, however, Presagis understood that to succeed with the scale and complexity of this project, an automated approach was needed. The robustness of the V5D workflow dramatically reduced delivery time and enabled rapid and automated content creation. Additionally, many of the customer’s existing workflows have been reused, eliminating the need to create new procedures “from scratch” to generate a new database in CDB format. By using V5D, the client received content generated in hours instead of weeks or months.

Case Study B: Country-wide rapid synthetic environment

Presagis was commissioned by a European Ministry of Defense to process data and generate a 60,000 km synthetic environment2 European country. A GeoINT initiative of government officials requested an assessment of the multiple benefits of V5D. The requirement was to complete the processing and generation of a large-scale virtual environment within 72 hours.

V5D has demonstrated the ability to quickly ingest new GIS data sources and create content not only for GeoINT, but also for generating analysis in emergency management assessment and response.

The Presagis mandate included two cases:

  1. A. Creation of a CDB from available open source data (Imagery, Elevation, Trees and Buildings OSM)
  2. B. Creating a CDB using AI/ML algorithms (from provided satellite imagery) to extract building footprints from provided satellite imagery. This case is usually when the vector data has not been updated and requires information extraction.

Each case was completed in less time than required by the original mandate (72 hours), and with incredible results. V5D was able to generate the following for each case:

Case 1: CDB in 7.5 hours – Includes source data ingestion, processing, and CDB creation.

Case 2: Total CDB 68 hours – Includes source data ingestion, extraction of building footprints from satellite imagery using AI/ML, process and creation of CDB. The V5D AI/ML algorithm was able to extract more than 100,000 buildings against the Case 1 OSM data.

Using AI/ML feature extraction, V5D was able to extract approximately 100,000 more building footprints than Case 1, which used OSM data as a source. Although the number of additional buildings may seem small for an entire country, it is significant considering that the data is almost 8 years old. This means that the number of new buildings that could be extracted using more up-to-date images would be significantly increased.

By combining the best tools with a next-generation architecture, V5D offers impressive performance in managing location-based content. From 2D to 3D, users can seamlessly transform diverse GIS data streams into rich, multi-spectral environments, complete with life patterns, humans, crowds, traffic, interactions, and behaviors. The result is a digital twin that offers more than surface, space and time – it offers the fifth dimension: context.

For more information, click here.

Sharon D. Cole