Aule SpaceはISROの施設で、カメラだけを用いた衛星ドッキング技術のTRL-6試験に成功。最終15メートルの自律近接運用を実証し、2027年の軌道上ドッキング実証と、衛星寿命延長用の「ジェットパック」衛星への搭載を目指す。

ISRO Spaceflight

𝗔𝘂𝗹𝗲 𝗦𝗽𝗮𝗰𝗲 𝘁𝗲𝘀𝘁𝗲𝗱 𝘁𝗵𝗲𝗶𝗿 𝗚𝗡𝗖 𝘀𝘁𝗮𝗰𝗸 𝗳𝗼𝗿 𝗱𝗼𝗰𝗸𝗶𝗻𝗴 𝗮𝘁 𝗜𝗦𝗥𝗢-𝗨𝗥𝗦𝗖!

Recently, Bengaluru-based Aule Space completed a closed-loop test of their GNC stack to be used for docking in orbit! 🛰🤝🛰

With their maiden orbital docking mission aimed for 2027, Aule Space is likely to become the 2nd company in the world to perform a 100% optical camera aided docking! 📸

While other spacecrafts (including ISRO's SpaDEx) uses expensive LiDAR & Laser Range Finder tech for determining the relative distance and orientation of the target spacecraft which they want to dock to, Aule Space's entirely camera-based computer vision architecture makes their system highly cost effective!

They plan to use this system on their 'jetpack' satellites which are essentially mission extension vehicles designed to service and extend the service life of satellites (especially in Geostationary Orbit) that are fully functional but have run out of propellants, by docking with them and acting as a jetpack, even if those satellites were never originally designed to be docked to.

As Aule Space is among the first to attempt docking like this using only cameras, there is a lack of a suffiently large image dataset containing close-up images of satellites in orbit to train their computer vision model on.

To tackle this, Aule Space set up their own dark room facility, used a light source to simulate the Sun, and moved a dummy satellite using robotic arms to simulate how satellites would appear and move in the darkness of space partially illuminated by sunlight to train their model.

Then using this model, they succesfully tested the final 15 meters of autonomous proximity operations at ISRO's Rendezvous Simulation Lab (where SpaDEx was tested) at ISITE in URSC! With this, their system achieved a technology readiness level of TRL-6!

Aule Spaceは、カメラのみで衛星ドッキングを行う技術のTRL-6試験に成功しました。
- ISROのランデブーシミュレーション研究所で、最終15メートルの自律近接運用を実証

- 暗室で太陽光と衛星の動きを再現し、画像データを収集して視覚認識モデルを訓練
- 2027年の軌道上ドッキング実証と、衛星の寿命を延ばす「ジェットパック」衛星への搭載を目指すぽい

インドのAule Spaceがカメラのみの自律衛星ドッキング技術を実証

関連するツイート