Patent Portfolio Overview

Patent portfolio: camera-based navigation safety technology

This portfolio is two granted US patents covering a camera-based safety system for autonomous vehicles and drones. Together they contain 33 claims across the dual-module safety architecture and a clear-passage-determining module for traffic-aware navigation decisions.

US Patent 12,001,207: dual-module camera-based navigation safety system

• Patent Number: US 12,001,207 — View full patent on Google Patents
• Issue Date: June 4, 2024
• Expiration: March 5, 2041
• Technology: Dual-module camera-based navigation safety system
• Assignee: Individual (self-owned)
• Patent Family: Application US16/987,612 (granted as US12,001,207, expires March 5, 2041); priority to NL2023628; related patents include EP3786756B1 (Europe) and continuation US12,530,030B2
• Technology Classification: G05D1/0088 (Autonomous decision-making), B60W60/001 (Autonomous road vehicle planning), G01C21/34 (Route guidance)

Patent abstract

US Patent 12,001,207 describes a dual-module safety system for controlling autonomous vehicles and drones. The system uses two distinct modules working together: a safety-determining module and a control module.

The safety-determining module captures live camera images, compares them to stored preprocessed reference images, and generates a safety value indicating whether navigation instructions can be safely executed. The control module only executes navigation instructions if the safety value exceeds a predetermined threshold. If the threshold is not met, the system requests human operator intervention. The system also accumulates and stores training data for future reference.

For complete legal text and detailed claims, refer to official USPTO records.

Key technical claims

1. Dual-module safety architecture

• Safety-determining module compares live camera images against stored reference images to calculate a safety value
• Control module only executes navigation instructions if the safety value exceeds a predetermined threshold
• Confidence-based safety thresholds with human operator handoff when uncertainties arise

2. Navigation decision logic

• Route-based navigation with safety value assessment before instruction execution
• Real-time path adjustment based on visual scene analysis
• Integration with machine learning models for predictive safety assessments

3. Autonomous system integration

• Designed for integration with autonomous vehicle control systems
• Camera-primary approach with optional acceleration sensor integration

Continuation patent: US 12,530,030

• Patent Number: US 12,530,030 — View full patent on Google Patents
• Issue Date: January 20, 2026
• Expiration: March 5, 2041 (terminal disclaimer, same expiration as US 12,001,207)
• Claims: 20 total: method claims (1-15), computer program product claims (16-18), system claims (19-20)

The continuation adds a clear-passage-determining module. This module receives live camera images and navigation instructions, then compares those images against stored images annotated with clear-passage values. Based on that comparison, it determines whether a given navigation instruction can be safely executed at that moment.

An example from the patent specification: the system learns that when the navigation instruction is "turn left at the next junction" and an oncoming vehicle has right-of-way, it should wait. Once subsequent camera images show the oncoming traffic has cleared, the instruction can proceed.

The continuation has 20 claims across three categories (method, computer program product, and system), compared to the original patent's 13 apparatus claims.

A terminal disclaimer ties both patents to the same expiration date of March 5, 2041.

View full continuation patent details

End-to-end neural network coverage

Claim 13 of US 12,530,030 covers "deep learning via network topology for converting navigation instructions into directional and acceleration values." The patent abstract describes the same pipeline at system level: converting "the navigation instruction(s) and the camera images into control values and acceleration values." Camera images in, control values out. That is the canonical end-to-end driving formulation.

The claims are implementation-agnostic. They cover this functional pipeline whether it is built with classical computer vision, convolutional neural networks, transformers, or a fully end-to-end learned model. The specification names "Nvidia Dave 2 network topology" as a suitable implementation, showing the patent contemplated end-to-end neural network driving from the filing date.

The clear-passage-determining module (Claims 7-11) uses what is now called behavioral cloning: the system learns from watching a human operator wait at a left turn with oncoming traffic, then applies that learned behavior in autonomous operation.

Claim 13 depends on Claim 1, so the deep learning conversion operates within the safety-gated control loop. The network produces control values only after the safety value exceeds the predetermined threshold.

For a full discussion of the end-to-end claims and how they apply to current industry architectures, see end-to-end neural network patent coverage.

Industry applications

The patents cover autonomous land vehicles and air vessels. The following are potential use cases inferred from the portfolio's technical scope, not explicitly stated in the patents.

Autonomous vehicles

• Consumer vehicles: safety systems for autonomous driving
• Commercial fleets: robotaxi and delivery vehicle applications
• Urban navigation: city driving with complex visual environments
• Highway systems: navigation with camera-based hazard detection

Drones and UAVs

• Delivery drones: navigation in populated areas
• Inspection drones: industrial facility navigation with obstacle avoidance
• Agricultural UAVs: autonomous crop monitoring with terrain safety
• Emergency response: search and rescue operations

AI navigation platforms

• Edge computing: real-time processing for autonomous systems
• Neural networks: training data accumulation for navigation AI models, end-to-end learned driving (Claim 13), and behavioral cloning from human demonstrations
• Sensor fusion: integration of camera data with other sensors
• Machine learning: camera-based image processing and safety assessment

Competitive landscape

The following analysis is inferred from the portfolio's technical scope, not statements in the patents themselves.

Market positioning

Several major technology companies are building autonomous navigation systems that rely on cameras. As the industry moves toward camera-primary approaches, this portfolio may be relevant to:

• Startups developing AV technology
• Automakers adding autonomous capabilities
• Technology companies entering the mobility market
• Drone manufacturers expanding into commercial applications

Strategic uses

These are business considerations, not claims made in the patents.

Freedom to operate: Licensing provides rights to use the patented technology in camera-based navigation systems.

Defensive IP: The portfolio can protect against competitor litigation and enable cross-licensing.

Credibility: Granted patents can be useful when talking to investors, customers, and regulators.

Licensing

This patent portfolio (US Patent 12,001,207 and US Patent 12,530,030) is available for licensing in the United States and United Kingdom. The two patents cover camera-based navigation safety assessment, dual-module control architecture, and clear-passage determination for autonomous land vehicles and air vessels.

Options include exclusive, non-exclusive, field-of-use, and geographic licensing. See our licensing page for details, or contact us to start a conversation.

Frequently asked questions

What does US Patent 12,001,207 cover?

US Patent 12,001,207 covers a dual-module camera-based navigation safety system for autonomous vehicles and drones. A safety-determining module compares live camera images to stored reference images and generates a safety value. A separate control module only executes navigation instructions if the safety value meets a predetermined threshold.

When do the patents expire?

Both expire on March 5, 2041, subject to maintenance fee payments. The continuation patent is bound to the same expiration through a terminal disclaimer. US 12,001,207 was granted June 4, 2024; US 12,530,030 was granted January 20, 2026. Both share a priority date of August 9, 2019.

What is the dual-module safety system?

It is the central mechanism in US Patent 12,001,207. The system separates autonomous navigation into two modules: (1) a safety-determining module that evaluates whether navigation instructions can be safely executed by comparing camera images against stored references, and (2) a control module that only carries out those instructions if the safety value exceeds a set threshold. If the threshold is not met, the system hands off to a remote human operator.

Who can license this portfolio?

Any company developing camera-based autonomous navigation for vehicles, drones, or other platforms. Licensing is currently available for the United States and United Kingdom. Options include exclusive, non-exclusive, field-of-use, and geographic arrangements.

What industries do these patents apply to?

Autonomous vehicles (consumer, commercial, trucking), drones and UAVs (delivery, inspection, agriculture), and AI navigation platforms. Any system that navigates using cameras, runs confidence checks against stored images, or hands off to a human operator when confidence drops may fall within the portfolio's scope.

Do the patents cover end-to-end neural network driving?

Yes. Claim 13 of US 12,530,030 explicitly covers "deep learning via network topology for converting navigation instructions into directional and acceleration values." The patent abstract uses the same language at system level: converting camera images and navigation instructions into control and acceleration values. The specification names Nvidia's DAVE-2 end-to-end driving network as a suitable implementation. The claims are implementation-agnostic, so they apply whether the underlying model is a CNN, transformer, or any other architecture. Claim 13 depends on Claim 1, meaning the deep learning conversion operates within the safety-gated control loop.

What does the continuation patent (US 12,530,030) add?

It adds a clear-passage-determining module. This module evaluates whether a navigation instruction can be executed given current traffic conditions by comparing live camera images against stored images annotated with clear-passage values. For example, it can determine that a left turn should wait until oncoming traffic with right-of-way has cleared. The continuation has 20 claims across method, computer program product, and system categories.

What is a terminal disclaimer?

A terminal disclaimer ties a continuation patent's expiration to the original patent's expiration. Here, both US 12,001,207 and US 12,530,030 expire on March 5, 2041, regardless of the continuation's later grant date. It also means both patents must be owned by the same entity.

How many claims does the portfolio include?

33 total. The original patent (US 12,001,207) has 13 claims covering the dual-module safety architecture. The continuation (US 12,530,030) adds 20 claims: method claims (1-15), computer program product claims (16-18), and system claims (19-20).