Patent Pending · USPTO 2026

Autonomous drone security with
temporal intelligence

FirmaChristo builds edge-AI autonomous UAV systems that see, remember, and reason about security threats over time — entirely onboard, with zero cloud dependency. Our patent-pending temporal reasoning engine detects what single-frame systems cannot.

Get in Touch → See the Technology

// Core Technology

Dual-Pipeline Temporal Reasoning

Our patent-pending architecture fuses real-time object detection with vision-language model reasoning to build persistent temporal scene understanding — directly on the drone.

Real-Time Object Detection

Lightweight CNN processes every frame at 5-30+ FPS, producing bounding boxes with spatial coordinates, classification, and appearance descriptors for persistent entity tracking.

Pipeline A · Every Frame

Vision-Language Reasoning

Quantized VLM performs chain-of-thought analysis on keyframes, producing structured threat assessments with full reasoning chains. Dynamically accepts operator BOLO search descriptions.

Pipeline B · Motion-Triggered

Persistent Scene Memory

Local relational database maintains a temporal graph of every observed entity across frames and patrol missions. Enables cross-patrol re-identification and historical anomaly detection.

Entity Tracking · Cross-Mission

Temporal Reasoning Engine

Evaluates configurable security rules against accumulated scene memory: loitering, new/missing objects, no-exit detection, zone violations, and tailgating — anomalies invisible to single-frame analysis.

Patent Pending · 8 Claims

Closed-Loop Actuation

Security alerts translate directly into UAV flight commands — position hold, orbital surveillance, reduced patrol speed — without human-in-the-loop latency. The drone responds autonomously.

Autonomous Response

// What We're Building

Products & Initiatives

From autonomous perimeter security to high-speed drone racing, our edge-AI stack is built to adapt.

Baldwin Security Drone

Autonomous perimeter security UAV running the full temporal reasoning stack on NVIDIA Jetson edge hardware. Dual-pipeline perception, persistent entity tracking, encrypted comms, and closed-loop physical response.

Patent Pending

AI Grand Prix — DCL 2026

Competing in the inaugural AI Grand Prix autonomous drone racing competition. Custom BitTemporal GateNet perception engine with temporal fusion, future prediction, and adaptive control — derived from our core IP.

In Development

Edge-AI Software Stack

Modular ROS2-based software platform: real-time detection, VLM reasoning, temporal engine, evidence recording, secured dashboard bridge, and operator BOLO injection — deployable on NVIDIA Jetson and equivalent hardware.

Deployed & Tested

Edge-Native Security Architecture

All perception, reasoning, and actuation occur onboard — no surveillance data ever leaves the drone. Minimal attack surface by design: no cloud servers to breach, no video streams to intercept. Encrypted operator links with API key authentication. Autonomous operation continues even under communications denial.

Architecture Defined

// Team

Who We Are

A focused engineering team building at the intersection of autonomous systems, edge AI, and security.

Jacob Florio

Founder & Lead Engineer

Electrical Engineering student with project experience in liquid rocket engines, CubeSats, biomedical research, and AI drone systems. Architect of the patent-pending temporal reasoning system. Proficient in C, C++, Python, ROS2, NVIDIA Jetson, and Isaac Sim.

Engineering

ML & Systems

Machine learning engineering expertise in neural network deployment, model optimization, and edge hardware integration. Experience deploying production ML systems at scale.

Business Development

Strategy & Growth

Go-to-market strategy, pilot customer engagement, and strategic positioning for defense and commercial security markets. Focused on enterprise security director buyer persona.

Autonomous drone security with temporal intelligence