The great power competition has returned, with Russia and China simultaneously advancing warfighting capabilities with increased lethality, range, and speed. The potential for lost military advantage by the United States is significant and military leaders at the Pentagon are pushing modernization initiatives to meet the demands of the National Defense Strategy (NDS).
Intelligence, surveillance, and reconnaissance (ISR) assets are critical to US Military dominance. ISR systems range in size from hand-held devices to satellites. ISR systems collect and process insights from unstructured data. These insights are used to support lighting fast information transfer for making more informed decisions in the field.
The Department of Defense (DOD) is positioning the U.S. and allied forces to outthink, outpace, and outmaneuver their adversaries, and ISR capabilities play a central role in these efforts. They plan to connect ISR sensors across all warfighting domains (space, air, land, sea, and cyber) directly with commanders and weapon systems, sharing data at an accelerated speed.

ISR is producing more data than can be reviewed manually
The DOD is inundated with exponentially growing ISR data from multiple sources and sees major potential to gain actionable intelligence from these data sources. The military services generating this data need to securely deliver the data to weapons, weapon systems, and commanders. The workforce required to process this data manually would be enormous and would struggle to execute their missions in competition and combat at a pace greater than the enemy.
To meet the demands of the new global strategic environment, the DOD intends to shift from a manpower-intensive force optimized for operations within a permissive environment to an automated and AI-enabled force capable of defeating peer adversaries within contested environments.
AI delivers insights from ISR faster
The goal is to process unstructured data at scale and turn around insights from downstream sources without the need to send the information to a command center for processing. AI automation speeds video footage review by many orders of magnitude and reduces the number of analysts needed for a given mission objective. AI shortens the time to get actionable intelligence for quick decision-making in the field. Computer Vision and AI platforms play a key role in ISR's ability to classify, detect, and track objects in images and video.
AI will improve situational awareness, accelerate decision-making, and reliably find, fix, and target elusive targets deep within enemy territory. Key benefits of AI include:
- Reduce human oversight and potential errors in video footage review and other incoming data.
Exploit intelligence information faster to make better life-saving decisions.
- Improve domain awareness for military perimeter surveillance, maritime compliance, and border security and tracking.
- Support ISR military projects faster by leveraging spatial, high altitude, aerial and terrestrial data in real-time.
- Make field decisions quicker by implementing Computer Vision into ISR infrastructure to gain insights and make predictions in near real-time.
- Stay ahead of changing field conditions even after the project is launched by using active learning and making your models smarter over time.
How AI accelerates the analysis of unstructured data
AI technologies leverage data, algorithms, computer power, and networked operations to ensure military readiness. AI can uncover insights from unstructured ISR data in the form of images, text, and full-motion video data to support faster field decisions by enabling analysts to categorize, search, sort, and filter their data. Key services provided by AI include:
- Classification - AI can analyze unstructured data streams in near real-time, and classify this data based on human-understandable concepts. Classification algorithms can be used on images, video, text, and audio data. Analysts simply input the data type of choice into a classification algorithm and AI will categorize the data based on pre-defined criteria.
- Object detection - Similar to classification, AI can identify any visually distinct object in images and videos.
- Object tracking - When working with objects in video data, AI can track unique objects over time.
Reverse image search - AI makes it possible to use images as inputs in search. Instead of searching based on keywords, analysts can search for images based on visual similarity to each other.
- Sentiment analysis - AI can analyze sentiment in multiple languages and provide instant high-level summaries of text passages.
- Data preparation - AI learns from data, and AI can actually help to pre-process the data that it learns from. AI-powered tools are essential to cleaning, labeling, and organizing data that is then used to train models to achieve specific mission objectives.