VANCOUVER, B.C. — In a technical release on Friday, February 13, 2026, SPARC AI Inc. (CSE: SPAI) announced a significant upgrade to its Overwatch platform, introducing machine learning (ML) models designed to continuously calibrate and correct drone telemetry data in real-time. The software-only solution aims to resolve targeting and navigation drift—a common failure point in low-cost sensors—without requiring hardware modifications or increasing unit weight.
The upgrade positions SPARC AI as a critical “intelligence layer” between commodity Inertial Measurement Units (IMUs) and mission-critical navigation outcomes. By analyzing the unique behavior of an individual drone’s telemetry stream during a short calibration flight, Overwatch identifies specific bias patterns and applies ongoing corrections during operational use.
Technical Optimization and Software-as-a-Sensor
The Overwatch platform utilizes a feedback loop where each connected flight enhances the system’s predictive accuracy. Key technical features of the upgraded release include:
- Telemetry Bias Correction: Uses AI priors to tighten navigation drift, enabling “commodity-grade” sensors to perform with the precision of high-grade, expensive hardware.
- Manufacturer Agnostic: Designed to operate across diverse drone manufacturers and varying IMU configurations.
- Operational Learning: Statistical confidence improves as the dataset grows across different flight regimes, altitudes, and sensor profiles, accelerating the onboarding of new fleet assets.
- Zero Signature Targeting: Maintains SPARC AI’s core capability of identifying target geolocation without the use of GPS, radar, lidar, or active laser range finders.
Alignment with “Drone Dominance” Initiative
The software-centric approach aligns with the U.S. Department of Defense’s Drone Dominance initiative. Launched in mid-2025 by Secretary of Defense Pete Hegseth, the program aims to field over 300,000 low-cost drones by 2027.
SPARC AI’s upgrade directly addresses a primary bottleneck of the initiative: improving the lethality and precision of inexpensive, mass-produced systems. By solving for accuracy through software, the Company preserves the range, flight time, and unit economics essential for large-scale, one-way attack drone deployments.
Executive Commentary
“With nearly every soldier now carrying a mobile device, the opportunity to deploy SPARC AI on phones is on the same order of magnitude as our drone opportunity,” said Anoosh Manzoori, CEO of SPARC AI. “Overwatch learns the behavior of each drone’s telemetry stream and applies correction to reduce targeting and navigation drift, improving accuracy and reliability without requiring new hardware.”
Expansion and North American Presence
Alongside the software rollout, SPARC AI has incorporated a dedicated United States subsidiary and commenced recruitment for key technical and strategic roles. This expansion is intended to accelerate engagements with North American defense contractors and support the deployment of SPARC AI on tactical devices, such as those provided by reseller Precision Technical Defence.