# Tactical-Edge Intelligence
**Entity class:** Distributed intelligence-processing architecture
**Domain:** ISR / edge processing / GEOINT / command and control
**Maturity:** Developed
## Definition
**Tactical-edge intelligence** is the processing and interpretation of sensor data near the point of collection and operational use rather than after transmission to a distant analytic center. Moving compute toward the sensor can reduce latency, conserve bandwidth, support disconnected or contested operations, and give operators locally relevant classifications or alerts in near real time.
## Functional Role
The architecture is especially important for wide-area imagery, video, radar, electronic signals, and autonomous or semi-autonomous platforms. AI and machine learning can prioritize frames, detect objects, compare spectral signatures, classify possible threats, and expose camouflage or decoys before the full dataset reaches a central system. Selected results can then enter [[wiki/GEOINT|GEOINT]], [[wiki/Multi-INT Fusion|multi-INT fusion]], and a [[wiki/Common Operational Picture|common operational picture]].
Tactical-edge processing does not eliminate central analysis. Edge systems operate under limits in compute, power, connectivity, model coverage, and contextual knowledge. A reliable architecture records model version, confidence, sensor conditions, and the path by which an edge inference becomes an operational decision.
## Counterterrorism Relevance
In the broader counterterrorism lineage, tactical-edge intelligence extends the effort to compress the interval between observation and authorized response. Its contribution is temporal and architectural: computation accompanies the sensor, so recognition can occur closer to collection and feed the [[wiki/Sensor-to-Decision Loop|sensor-to-decision loop]] sooner.
## Relationships
- **master collection:** [[collections/Terrorism, Counterterrorism, and the Intelligence Environment|Terrorism, Counterterrorism, and the Intelligence Environment]].
- **intelligence discipline:** [[wiki/GEOINT|GEOINT]] and [[wiki/Pattern-of-Life Analysis|Pattern-of-Life Analysis]].
- **sensor layer:** [[wiki/Sensor Fusion|Sensor Fusion]] and [[wiki/Real-Time Observability|Real-Time Observability]].
- **operational loop:** [[wiki/Sensor-to-Decision Loop|Sensor-to-Decision Loop]].
- **command architecture:** [[wiki/CJADC2|CJADC2]] and [[wiki/Command and Control Systems|Command and Control Systems]].
- **tempo:** [[wiki/Machine-Speed Intelligence|Machine-Speed Intelligence]].
## Sources / Provenance
- [RTX — Raytheon Awarded Contract for Poland Airborne Reconnaissance System, January 28, 2026](https://www.rtx.com/news/news-center/2026/01/28/rtxs-raytheon-awarded-197-million-contract-for-poland-airborne-reconnaissance-s)
**As of:** 2026-09-23
## Simple Reminders, Quotations, and Thoughts
> "The MS-110 system brings advanced capability by pushing next-generation processing to the tactical edge to defeat camouflage and decoys in near real time."
> **— Dan Theisen**, *January 28, 2026, RTX MS-110 announcement*
[[reminders/Information/Push Processing to the Tactical Edge by Dan Theisen|Push Processing to the Tactical Edge by Dan Theisen]]
> "Defense and military organizations can wield the power of LLMs and cutting-edge AI - all on their network, from classified systems to devices on the tactical edge."
> **— Palantir**, *accessed September 23, 2026, AIP for Defense*
[[reminders/Information/Defense AI Can Reach From Classified Systems to the Tactical Edge by Palantir|Defense AI Can Reach From Classified Systems to the Tactical Edge by Palantir]]
> "Palantir is deploying Maven Smart System across the entire Department of War, moving cutting-edge AI from the lab to America's warfighters at speed and scale."
> **— Palantir**, *Q1 2026, Q1 2026 Business Update*
[[reminders/Information/Maven Moves AI to Warfighters at Speed and Scale by Palantir|Maven Moves AI to Warfighters at Speed and Scale by Palantir]]
> "To enhance resilience and mitigate risk, we design collection and processing systems that move analytics closer to the source—reducing reliance on vulnerable central hubs."
> **— Booz Allen**, *accessed September 23, 2026, Enhancing Intelligence Operations with AI*
[[reminders/Information/Move Intelligence Analytics Closer to the Source by Booz Allen|Move Intelligence Analytics Closer to the Source by Booz Allen]]
> "GAMBLER brings forth the ability to create new models in the field for automated imagery processing—it’s hours or days, instead of months, to train the system on a unique model which is specific to warfighter needs."
> **— Frank Whitworth**, *May 19, 2025, 2025 GEOINT Symposium Keynote*
[[reminders/Information/Edge AI Models Can Be Trained in Hours Instead of Months by Frank Whitworth|Edge AI Models Can Be Trained in Hours Instead of Months by Frank Whitworth]]
> "Designed to operate at the tactical edge, these platforms leverage machine learning to detect and classify threats faster, prioritize sensor tasking dynamically, and support real-time targeting decisions."
> **— CACI**, *accessed September 23, 2026, ISR/EW Mission Management and Data Analytics*
[[reminders/Surveillance/Machine Learning Can Detect and Classify Threats at the Tactical Edge by CACI|Machine Learning Can Detect and Classify Threats at the Tactical Edge by CACI]]
> "Edge Computing Devices: Man-packable and low Size, Weight, and Power (SWaP) technologies that permit computing devices at the edge that for processing of AI/ML algorithms such as Natural Language Processing, General Information Processing, Video/Graphics Processing without any back-end server connectivity."
> **— USSOCOM**, *accessed September 23, 2026, USSOCOM Capability Areas of Interest*
[[reminders/Information/Edge Devices Can Run AI Without Back-End Servers by USSOCOM|Edge Devices Can Run AI Without Back-End Servers by USSOCOM]]
> "CADS' ability to quickly process data and run third-party algorithms that prioritize threats, with almost no latency will significantly enhance survivability for military personnel."
> **— Bryan Rosselli**, *February 24, 2025, AI/ML-Powered Radar Warning Receiver Demonstration*
[[reminders/Surveillance/AI at the Sensor Can Prioritize Threats With Almost No Latency by Bryan Rosselli|AI at the Sensor Can Prioritize Threats With Almost No Latency by Bryan Rosselli]]