Using NVIDIA Jetson for AI-Powered Repair Robots

Goran Vuksic takes his robot to technology conferences across Europe, showcasing the powerful capabilities of edge AI platforms with this familiar “Star Wars” character.

Goran Vuksic leads a project to create a repair robot (Pit Droid) from “Star Wars” in the real world. The “Star Wars” series is beloved by many, and the repair robots in the films are responsible for fixing and maintaining the Podracers that traverse the series.

This edge AI “Jedi” uses NVIDIA Jetson Orin Nano Developer Kit as the brain of the robot, enabling this robot, which stands less than four feet tall and has only a simple webcam for an eye, to recognize objects and turn its head towards them.

Vuksic, who is from Croatia, currently works in Malmö, Sweden. Recently, he traveled with his repair robot through Belgium and the Netherlands, attending several technology conferences where he introduced hundreds of people to computer vision and AI, performing engaging live demonstrations with the robot.

The repair robot sees the world for the first time

Vuksic describes himself as a “Star Wars” fanatic. He is now an engineering manager at a company in Copenhagen and upgrades the robot’s capabilities in his spare time. He is also a member of the NVIDIA Startup Accelerator Program and co-founder and CTO of syntheticAIdata.

The company creates visual AI models using affordable synthetic data and builds and runs 3D tools and applications via a NVIDIA Omniverse connector.

Vuksic has been recognized as a Jetson AI expert by NVIDIA and named a “Most Valuable Professional” in AI by Microsoft. He began his journey into AI and IT about ten years ago while working at a startup that classified tattoos using visual AI.

Since then, he has held various positions, including engineering and technical manager, developing IT strategies and solutions for multiple companies.

Being a super sci-fi fan since childhood, Vuksic has always had a keen interest in robotics.

Vuksic has also joined the NVIDIA Developer Program. He mentioned, “After watching ‘Star Wars’ and similar films, I started to imagine how robots could see and do things in the real world.”

Today, he is realizing this vision through the repair robot project that uses the NVIDIA Jetson platform. He has been using this platform since the first NVIDIA Jetson product was released ten years ago.

Vuksic reading for the repair robot

Besides tinkering with computers and robots, Vuksic enjoys playing bass guitar with friends in a band.

Vuksic builds the repair robot for both entertainment and education.

He often speaks at tech conferences, bringing the repair robot on stage to interact with the audience, demonstrating how it works, and encouraging others to build similar projects.

Vuksic, his startup co-founder Sherry List, and the repair robot

at the Techorama conference held in Antwerp, Belgium

Vuksic states, “We live in a connected world where everything is exchanging data and becoming increasingly automated. This is very exciting, and there may be more robots in the future to assist humans in tasks.”

With the NVIDIA Jetson platform, Vuksic stands at the forefront of robotic innovation alongside a developer ecosystem leveraging edge AI.

Vuksic’s repair robot project took four months, during which he printed the robot’s body parts using a 3D printer and assembled them.

He then installed the Jetson Orin Nano Developer Kit as the brain of the robot’s head, which can turn in all directions using two motors.

Vuksic installs the NVIDIA Jetson Orin Nano

Developer Kit in the head of the repair robot

Vuksic states, “Jetson Orin Nano can process camera images in real-time. It’s only the size of the robot’s head but has such powerful processing capabilities, which is truly amazing.”

He also uses Microsoft Azure to process data in the cloud for object detection training.

“The best part of this project is connecting the robot with Jetson Orin Nano, allowing easy AI operation for the robot to move based on what it sees.” To enable others to attempt building robots, Vuksic wrote a step-by-step technical guide on how to build the robot (https://www.hackster.io/gvuksic/nvidia-jetson-orin-nano-powered-pit-droid-7da0e8).”

“The hardest part is traveling with the robot. Every time I go through security, I have to open the bag with the robot parts and explain: this is just my big toy!”

Learn more about the NVIDIA Jetson platform here:

https://www.nvidia.cn/autonomous-machines/embedded-systems/