Detect
Analyze camera feeds for a potential rifle, pistol, or knife, then create a time-stamped event tied to the facility and camera.
- Parallel camera analysis
- Wide and zoomed event images
- Confidence-based event timeline
Public-safety research & development
Soul Shield is developing a connected safety system that combines computer vision, human verification, and rapid mobile alerts—designed to help facilities gain the time to act.
The mission
In a fast-moving emergency, every moment between detection and coordinated action matters.
Soul Shield is a Nevada nonprofit advancing practical technology for safer facilities. The Mass Defense project connects existing cameras, computer vision, distributed human review, and mobile messaging into one coordinated system.
The goal is focused: identify a potential weapon, verify the threat, and alert the people who can respond—without removing human judgment from the decision.
“Save lives through weapon detection technology and safer facilities.”
The Mass Defense system
The first two stages are in active testing. Later stages remain research concepts intended to add communication, tracking, and intervention capabilities as the project advances.
Analyze camera feeds for a potential rifle, pistol, or knife, then create a time-stamped event tied to the facility and camera.
Send detections to distributed human examiners. When a threat is confirmed, notify designated responders with facility context.
Explore two-way audio and mobile camera platforms that could support communication, tracking, and distraction while help is on the way.
Investigate robotic systems that could eventually help create distance, reduce exposure, and support physical intervention.
Evidence over promises
Soul Shield has progressed from a single-device classifier to a multi-camera detection and facility-awareness system. Each test has produced a clearer next step.
The fourth system test brought six security cameras into one live workflow, adding a chronological event timeline, paired wide and zoomed images, and integrated facility tracking.
Alpha One
First classifier test on an iPhone, paired with SMS messaging and a basic command center.
Watch on YouTubeAlpha Two
Introduced security-camera video, new machine-learning models, and detection across three weapon classes.
Watch on YouTubeAlpha Three
Added fourth-generation models, confidence filtering, date stamps, and side-by-side wide and zoomed views.
Watch testBeta One
Expanded to six cameras and introduced an integrated mini-map with dynamic location tracking.
View latest testAlpha Three introduced upgraded computer vision, zoomed weapon views, confidence filtering, and the ability to test from recorded video.
Technical view
Designed to analyze multiple camera feeds and organize detections into a reviewable event stream.
Places human confirmation between automated detection and the responder alert.
Concept work for two-way audio, remote communication, and mobile camera tracking. These capabilities are not presented as deployed production features.
Long-term exploration of human-supervised robotics intended to reduce responder exposure and support safer physical intervention.
Henderson, Nevada
The people behind the work
Brian Adams leads the Mass Defense project, bringing a software background and specialization in computer vision to a deeply human problem.
Soul Shield is organized as a Nevada nonprofit. The project is seeking grant support to expand technical development, bring in specialized expertise, acquire equipment and services, and move beyond a single test location.
Help move the work forward
Soul Shield is looking for pilot facilities, technical collaborators, and grant partners who believe prevention technology should be tested carefully, transparently, and in the real world.