The current state of international affairs and conflicts in many parts of the world requires a constant state of vigilance and readiness for all US defense and intelligence operations. Part of this preparation includes a deep and accurate understanding of the ever-evolving geospatial landscape, as close to real-time as possible.
Geospatial knowledge has always been a critical component of military intelligence, aiming to minimize the time window of data capture from analysis to dissemination. Today this time lag is widening as the volume, variety and velocity of data increases at an unprecedented rate. More information ironically leads to slower (and less accurate) decision making. At the highest level, this delay can have a negative impact on the geographic intelligence required for any type of military operation.
There are several types of data that are often held in different databases – including LiDAR (light detection and ranging), multispectral, hyperspectral, SAR (synthetic aperture radar), aerial and other modalities. The high cost, inflexibility, and inability to share geospatial data between systems continues to be a major barrier because geospatial analysts are forced to use inefficient, time-consuming, and error-prone data transfer methods.
Any single data set provides only a limited amount of information and can only be used for a limited number of purposes. Integrating and linking multiple data sets or generated data products (related to a specific area of interest) is essential to gain more information, answer additional questions, and make better decisions. In fact, geospatial data is most useful when multiple formats are shared, analyzed, and used in conjunction with each other.
For example, an American professor recently received the prestigious Fulbright US Scholar Award for his geospatial research focused on humanitarian aid. This project involves synthesizing a huge amount of geospatial data, including mobile phone GPS tracking, remote sensing images, location references in text and more, to better understand and address the movement of Ukrainian refugees in Poland.
However, a data analyst must download multiple file formats and develop their own processing pipelines in order to synthesize and enrich data. Before starting a processing task, an analyst must search multiple databases to find the data he needs, and then download that complex data in multiple formats as inputs to a processing pipeline, each input requiring its own API . In a defense example, target detection using hyperspectral data requires a custom processing pipeline that also incorporates aerial imagery for the environment and possibly point clouds for advanced 3D visualization.
This approach limits the possibility of rapid processing across sources. There is no single place for all geospatial analytics and machine learning – including point cloud mining, image processing, feature detection, change detection, or support for digital twins – which prevents deeper contextual understanding.
Rapid processing from multiple sources is key to achieving the type of integrated richness that supports faster evaluations. Based on the defense target example above, beyond accessing and capturing basic data, this type of analysis adds another layer of complexity because there are different closed-source and open-source tools for analyzing each different type of data. Currently, advanced image analytics require custom tools with limited API integration. Imagine if there was a single API that optimized data access and could be integrated into these tools.
Finally, today’s geospatial analysts face limiting computational limitations. In particular, geospatial analysts often have to spin up clusters, which slows down the time to acquire information and limits the ability to parallelize operations. Advances in serverless architectures eliminate this need, allowing developers to easily spin up and download applications without worrying about or having to wait for hardware access.
We need a better approach, one that delivers information in minutes rather than days and is achieved through:
— A single platform to support all data modalities – there must be an efficient and unified method of storing and analyzing all geospatial data and derived results.
— Distributed and highly scalable compute – enabling geospatial analysts to fully embrace the cloud to run any pipeline at scale without having to launch and fire up clusters. and
— All this must be achieved while simultaneously protecting sensitive information and ensuring data integrity. Compliant and isolated on-premises capabilities should be in place to meet data sovereignty requirements for both the mission and your partners.
Geospatial knowledge continues to provide a vast repository of knowledge that can be used to improve defense and intelligence operations, and at a higher level, human society. However, the volume, variety and speed of this massive data requires a new approach to coherent management, as current approaches are too fragmented.
Doing so will be key to maximizing the power of geospatial information in the coming years, hopefully turning data into life-changing intelligence within increasingly tight timeframes.
Norman Barker is the vice president of Geospatial at TiledB. Prior to joining TileDB, Norman focused on spatial indexing and image processing and held engineering positions at Cloudant, IBM and Mapbox.
Read the original at Defence247.gr