AI-AV
Vision-Based Target-Oriented Adaptive Control
AI-AV provides vision-based guidance for autonomous platforms. The system uses camera input to identify, track and pursue targets—designed to operate without GPS for the relevant guidance function.
What AI-AV Does
AI-AV is a guidance system that uses visual input to control autonomous platform movement toward a target. The system processes camera data in real-time, measures target position relative to the platform, and generates guidance commands.
The architecture is designed to operate without reliance on GPS or external positioning systems for the guidance function. This makes it suitable for environments where GPS is unavailable, unreliable, or contested.
Capabilities
Core Functions
Visual Target Pursuit
Identifies, tracks and pursues targets using onboard camera and AI processing. Maintains continuous lock within field of view.
Designed for GPS-Denied Operation
Architecture designed to operate without GPS for the relevant guidance function. Subject to platform and sensor configuration.
Adaptive Guidance
Dynamically adjusts heading and speed to compensate for target movement. Responds to changing conditions in real-time.
Platform Integration
Integrates with compatible platforms through defined interfaces. Specific integrations validated for supported configurations.
Platform Support
Integration Domains
Aerial
Fixed-wing and multirotor UAVs with compatible interfaces
Ground
Unmanned ground vehicles with appropriate sensor mounts
Marine
Surface and underwater vehicles with vision capability
Platform-agnostic is an architectural objective. Specific integrations are validated for supported configurations and interfaces.
Approach
Key Differentiators
Designed for GPS-Denied
Vision-based guidance is the primary navigation method—not a fallback when GPS fails. The architecture is designed from the ground up to operate without satellite navigation for the relevant guidance function.
Deterministic Control
Predictable, testable, explainable control logic. Critical for safety certification and mission planning. Behaviour can be validated before deployment.
Lightweight & Efficient
Designed to run on modest embedded compute. No GPU required for core guidance function. Suitable for SWaP-constrained platforms.
Sensor Flexibility
Compatible with RGB, thermal, and low-light imaging sensors. Effectiveness varies with sensor quality and environmental conditions.
Intellectual Property
AI-AV technology is protected by filed patent applications. The RLK Group maintains original, protected technology developed in-house.
- Filed patent applications covering core guidance methodology
- RLK-developed IP—not a third-party integration or dependency
Note: "Filed patent application" status. Patent applications protect claimed subject matter, not every marketing statement made about a product.
Integration & Testing
AI-AV integration requires:
- Compatible platform with defined control interfaces
- Suitable visual sensor with adequate resolution and frame rate
- Embedded compute platform meeting minimum requirements
- Defined test environment for validation
Performance varies with platform characteristics, sensor quality, and environmental conditions. Specific performance claims require validation in the target configuration.
Discuss an Integration
Tell us about your platform, interface requirements, and the guidance capability you need. We will scope an integration approach and define test boundaries.