All Domain Geospatial Object
Tracking, ML-Based Detection,
and Alerting

ATA Provides Global High-Velocity
Object Tracking

Cloud-native, container-based technology that collects, processes, and analyzes geospatial readings of moving objects to support decision making informed by real time data and analytics.

Built on ATA’s Jetstream streaming analytics platform, the geospatial data processing and analytics solution uses advanced machine learning (ML), environmental modeling, and computer vision techniques to track unknown objects through geospatial readings, detect and alert on anomalous or restricted object behavior, and predict object behavior to inform an operational response – all within a modeled geolocation within space, air, land, or sea.

The solution has capabilities in multi-sensor data integration, ML performance monitoring and tuning, high velocity data pipelining and processing, and interactive real-time dashboard visualizations.

It utilizes ML to correlate readings to track objects and learn the expected behavior in terms of direction, velocity, location, and altitude (among other attributes) in a given geolocation. It cross references this with discrete rules and models that represent restricted spaces to determine objects that are anomalous or in violation of these rules and mapped and viewed in real time on web-based dashboards.

This solution is architected to be compatible with all major cloud providers or on-prem deployments, and is built on today’s leading open source technologies.

Problems Addressed Include:

The system can integrate with multiple streaming sensor data feeds that operate across multiple surveillance domains (air, space, land, sea), execute ML models against those datasets to track and identify potentially threatening objects, and present the data and analyses in dashboards or to downstream systems for real-time monitoring, alerting, and triaging.

An instance can be instantiated on-demand, with support for data flow configuration and ML tuning. Users can then securely access the data and view them through geospatial time-indexed dashboards.