DEVELOPEMNTS

Visual Inertial Odometry

GPS-free

Indoor / outdoor nav

Fusion

Camera + IMU

< 10 m

90th percentile error

TRL 5-6

Relevant env validated

Visual Inertial Odometry (VIO) enables drones and autonomous systems to estimate position and orientation in 3D space by fusing camera imagery with IMU data.

Unlike GPS-only navigation, VIO works both indoors and outdoors. It is designed for GPS-denied or electronically contested missions where continuity of autonomous flight is critical.

Rodella VIO integration on aerial platform

VIO Program Specs

System

VIGNETTE vision navigation engine

Input Sensors

Downward camera + IMU

Feature Stack

SuperPoint + LightGlue

Cold Start

< 3 sec heading resolution

Latency

150-400 ms per frame

Update Rate

5-10 Hz to EKF

Median Error

2.2 meters (PoC corridor)

90th Error

< 10 meters

Corridor Tested

~1.4 km

Compute

Jetson Orin Nano class

Added Weight

< 150 g target

Power

< 12 W target

Navigation Core

The VIO core continuously matches drone camera frames with geo-referenced satellite references, then fuses those fixes with inertial data for stable state estimation without GNSS.

VIO mission conditions in GNSS-denied environments

GNSS-Denied Operations

Designed for electronic warfare zones, dense urban canyons, and indoor missions where GNSS can be jammed, spoofed, or unavailable. VIO keeps autonomous flight functional in these denied environments.

Autopilot Integration

Position estimates are sent to ArduPilot/PX4-compatible EKF pipelines as an external vision source through MAVLink, enabling drop-in deployment without custom flight controller firmware changes.

VIO integrated with mission-ready platform