HomeElectronics NewsAI Powered Star Tracker Advances Satellite Navigation  

AI Powered Star Tracker Advances Satellite Navigation  

A satellite guidance system uses embedded AI and aims to improve orbital pointing accuracy under extreme radiation and signal noise conditions. How?

TM2Space to Develop StarSense under IN-SPACe's TAF, India's First AI-Powered Star Tracker
TM2Space to Develop StarSense under IN-SPACe’s TAF, India’s First AI-Powered Star Tracker

TakeMe2Space has introduced an AI powered star tracker, StarSense, designed to improve satellite attitude determination through onboard intelligence and optical precision. The system has been backed under IN SPACe’s Technology Adoption Fund and is being developed as a flight qualified subsystem for future satellite missions.

Star trackers are optical navigation instruments used to determine a spacecraft’s orientation by analysing star field patterns. They form a critical part of a satellite’s attitude determination and control system, where even minor errors can affect imaging accuracy, communication alignment, and mission stability. StarSense enhances this function by integrating embedded AI directly into the sensing and processing pipeline.

The main improvement comes from how star field data is processed. Instead of relying only on fixed geometric matching algorithms, the system uses an onboard AI accelerator to refine centroid detection and pattern recognition in real time. This helps maintain accuracy in challenging orbital conditions such as radiation induced noise, sensor interference, and partially obscured or dense star fields.

The system is being developed in two variants. A lightweight configuration targets CubeSats and academic missions with strict constraints on size, weight, and power. A higher precision version is designed for larger spacecraft requiring tighter pointing accuracy and long duration stability. Both variants use a shared optical architecture with radiation protected sensor design.

Its hardware stack combines an electro optical sensor, AI assisted processing, and shielding for the CMOS imaging unit. A hybrid AI and FPGA architecture is also used to reduce latency during spacecraft manoeuvres, enabling faster correction cycles in dynamic orbital conditions.

Ronak Kumar Samantray, Founder and CEO, TakeMe2Space, adds, “Owning the full hardware stack is not a luxury for what we are trying to build in orbit, it is a necessity. Every sovereign subsystem we develop today brings the vision of a large-scale orbital data centre, built and controlled in India, that much closer.” 

Click here to find the official announcement.

Saba Aafreen
Saba Aafreen
Saba Aafreen is a Tech Journalist at EFY who blends on-ground industrial experience with a growing focus on AI-driven technologies in the evolving electronic industries.

SHARE YOUR THOUGHTS & COMMENTS

EFY Prime

Unique DIY Projects

Electronics News

Truly Innovative Electronics

Latest DIY Videos

Electronics Components

Electronics Jobs

Calculators For Electronics