AI-powered Dynamic Targeting Technology, let’s explore it further!

What is it?

Dynamic Targeting is an onboard software system powered by AI that enables satellites to independently identify, assess, and react to phenomena on Earth instantly.

Created by: Created by NASA’s Jet Propulsion Laboratory (JPL) in partnership with Open Cosmos and Ubotica, through NASA’s Earth Science Technology Office.

Goal:

To enhance the quality and relevance of satellite data through the smart selection of observation targets such as clear-sky areas, wildfires, storms, and volcanic eruptions, without relying on ground-based commands.

Lookahead Sensor Method: The satellite leans forward to gather images up to 500 km ahead, assessing possible obstacles or goals.

AI Cloud Detection: An onboard AI system differentiates between cloud and clear areas by analyzing optical and near-infrared images.

Decision in less than 90 Seconds: According to the analysis, the satellite independently modifies its imaging strategy, recording solely clear-sky terrestrial information.

Pivot and Capture Mechanism: Upon evaluation, the system pivots to a nadir angle and records the enhanced observation frame in 60–90 seconds.

Main Characteristics:

Real-Time Autonomy: No terrestrial oversight required; satellite determines imaging target autonomously in space.

AI Edge Processing: Employs sophisticated on-device neural networks, minimizing communication lags.

Lightweight Execution: Evaluated with CubeSat-sized payloads, demonstrating adaptability to smaller satellites.

Cloud-Avoidance Optimization: Drastically enhances the usable imagery output, addressing a significant data-waste issue in Earth observation.

Scalability to Constellations: Can be incorporated into multi-satellite groups for synchronized observations, inclusive of planetary missions.

Strategic Importance:

Enhanced Climate Monitoring: Boosts tracking of infrequent, transient climate occurrences such as wildfires and ice storms.

Operational Efficiency: Decreases storage, processing, and downlink demand, maximizing satellite bandwidth.

Disaster Preparedness Instrument: Might independently identify volcanoes, hurricanes, and various other calamities in real time, supporting early alert systems.