Welcome to the Jetson Projects forum!

Please post about your Jetson based projects here! Full writeups are welcome, or just a few sentences with links to Hackaday, GitHub, or wherever you are documenting and sharing the project. Works in progress are welcome. Featured projects may be selected to highlight on the Jetson Projects page.

Please continue to use the other Jetson forums for community and NVIDIA support.

We look forward to seeing what you create!

1 Like

I’m looking forward to see what great projects people come up with. In the meantime I just wanted to say I put a JetBot through its paces today, following along with the setup and Collision Avoidance notebooks. All the instructions were incredibly well thought out and easy to follow. With only about 120 training images the bot was already doing a pretty-good job of wandering around our family room. The only weird thing I found is that the .ipynb files didn’t seem to want to download from Jupyter when I used Chrome. Switched to Edge and they were fine.

My favorite part is that the entire cycle of data collection, training (well, given a pre-trained AlexNet model to start with:)), and running can all be done entirely on the robot. Congrats on a great DIY learning project!

– David

Thanks David, great to hear you have been enjoying your JetBot!

I cannot find any simple example software in my Jetson Nano. It seems to be completely empty!!
Not even a help file… What a con!!

1 Like

You probably should post for the Nano in the Nano forums:

Briefly though, it is common to use JetPack/SDK Manager to install the extra components. Flash and “extras” are really separate steps. You’ll find base installs are more or less just Ubuntu Linux.

See this for listings of JetPack versions (you always use the same JetPack/SDKM version for adding packages as was used to create the root filesystem…a.k.a., “L4T” or “Linux for Tegra”. You may need to go there, log in, and go there again since redirect doesn’t work:

A listing of L4T versions is here (L4T is what goes in the Jetson when it is flashed…or what an image was created from):

You’ll notice the L4T page lists a Nano as compatible with R32.1 through R32.2.1. These in turn are flashed with JetPack/SDKM 4.2 through 4.2.2. JetPack/SDKM are really a front end tool running on a Linux Ubuntu PC, for flash or package additions (it may not be obvious when you start using the tool, but you can uncheck flash and leave only install of tools to the Nano if you are using the correct JetPack/SDKM version). Documents there too go with each release. The L4T and a given CUDA version are meant to go together and normally are not mixed. The image you may have used to add to an SD card was created with this tool. Normally image creation and tool addition are separate steps.

Good day,
I went to purchase the jet bot and found that it was discontinued. I am looking at a starting point to Nvidia and robotics. I have the nano, jetson tx2, and xvaier for my classroom. Any assistance in purchasing a kit/getting started would be appreciated.

Hi dmeyers99, where were you trying to purchase the JetBot? Was it a kit or from individual components?

If you check out the following page, various distributors for JetBot kits are listed:

Those distributors which are out of stock should have stock again shortly.

We’ve posted a bunch of handy resources and how-to guides from the community under the ‘Community Resources’ section found here:


Thanks to everyone who has contributed!

1 Like

Hello Team,
I am a Research student in AI and I have previous experience as Firmware engineer , more into embedded systems. I would like to know what extra AI related projects we can do after purchasing Jetbot, Facial Recognition, Line tracking, Obstacle Avoidance are already present as per of Nvidia library , which new interesting AI problems can be solved using the vision system of Jetbot, or Please list few projects and respective references for AI project ideas to be implemented using Jetbot.

Please do the needful.

Thanks and Regards
Abhijith N.M.

Hi @abimsstudy, you can find some drone project using Jetson on this page: https://developer.nvidia.com/embedded/community/jetson-projects

And also here: https://github.com/nvidia-ai-iot/redtail

Hello Team Nvidia,
Thank you for the quick response. The link you shared me definitely helps.
Please do share me projects references which are specific to Deep reinforcement learning as well. Please do the needful.

Thank you.

Abhijith N.M.

See these projects about reinforcement learning:

Hello Team,

I am checking for Jetbot in Nvidia site. we have.

  1. SparkFun JetBot AI Robot Kit

  2. Waveshare JetBot AI Kit

  3. Yahboom AI Robot for NVIDIA Jetson Nano B01 A02, Coding Robotics Kit with Autopilot, Object Tracking, Face and Color Recognition.

SparkFun JetBot AI Robot Kit has interactive AI tutorials, JetBot ROS (Robot Operating System), and out-of-the-box AWS cloud support.

Waveshare JetBot AI Kit has advantage of NVIDIA AI for facial recognition, obstacle avoidance, object tracking, and line following. even recharge the Waveshare JetBot while it’s running

Yahboom AI Robot for NVIDIA Jetson Nano B01 A02, Coding Robotics Kit has

  • Customized 8 million HD camera, 3 DOF camera head
  • 87 kinds of 178 structural components including customization
  • High-precision PID algorithm color tracking application capable of fast response
  • High-precision PID algorithm face tracking application capable of fast response
  • Obstacle avoidance model training multi-environment automatic obstacle avoidance application
  • Object-following application with high-precision differential PID algorithm that can respond quickly
  • Track model training automatic driving application
  • FPV (first person view) video real-time display independent of PC control, motion remote control,Multi-AI function integrated in one mobile phone mobile Android, IOS dual system-side APP blessing
  • PC side interactive programming through a web browser, providing basic motion control, AI framework use,AI model training, handle remote lever full range variable speed remote control, automatic obstacle avoidance, automatic follow, automatic driving, etc.The underlying code multi-routine tutorial, Jupyter notebook.
  • AI online programming debugging, AI model training
  • Support CUDA acceleration, PyTorch, TensorFlow and other mainstream AI frameworks and AI Getting Started Machine Learning Platform from 0 to 1 Tutorial

Please do let me know which AI kit is better considering I can execute the Nvidia example projects mentioned above.
And with best AI framework and AI library support and I can build cutting edge AI related projects.
Better build quality of robot body and better ROS community support.

Please do the needful.

Thanks and Regards
Abhijith N.M.

1 Like

All of the JetBot kits that run the project from https://github.com/NVIDIA-AI-IOT/jetbot are shown on this page:

I believe the Yahboom one is a different kit that uses different software than the NVIDIA JetBot project. So if you want to run JetBot, I recommend picking one of the JetBot kits from above. Which JetBot kit is up to your personal preferences and availability.

1 Like

Hello Team,

Which one would be better considering execution of Nvidia AI projects and Cutting Edge AI concepts.

  1. SparkFun JetBot AI Robot Kit

  2. Waveshare JetBot AI Kit

Thanks and Regards
Abhijith N.M.

The link to the Waveshare JetBot kit is here: https://www.amazon.com/Waveshare-JetBot-AI-Kit-Accessories/dp/B081LF98T8/?th=1

(what was previously linked to was just the accessories)

Both of the SparkFun and Waveshare kits run the NVIDIA JetBot project from GitHub, so their AI capabilities are roughly the same. At that point, it is basically up to which of the robot assemblies you prefer.

Since the SparkFun kit is currently on backorder, and Amazon has some stock of the Waveshare kit, perhaps that can help you decide.

Dear Sir:

Our company is new to the Nvida Xavier NX. I am hoping to share my project onto the Nvidia community project. May I ask how I can do? Could you please share more details to me? Your feedback is highly appreciated.


Hi @dusty_nv
this is my modest contribution to the war against the COVID19: I realized a Fever Control application on Jetson Nano integrating the FLIR Lepton3 thermal sensor.
The project is mainly a tutorial about using the SPI and I2C ports of the Nano at the same time, changing the buffer size of the Nano to match the requirements of the FLIR sensor.
Thermal images retrieved from the FLIR sensor are analyzed searching for temperatures in the Human Body range and triggering warnings and alerts when the values are too high.

Blog post: https://www.myzhar.com/blog/jetson-nano-with-flir-lepton3/
Hackster: https://www.hackster.io/Myzhar/covid19-fever-control-with-jetson-nano-and-flir-lepton3-7f57fc
YouTube: https://youtu.be/SFStaq--3-U


1 Like

Hi @dusty_nv
Here is my new interesting Jetson project called patrolling robot. The idea is to have a robot that patrols the huge parking space under or inside the commercial building to detect the illegal parking cases. As a prototype, I have integrated Jetbot (Jetson Nano inside) and 2 CSI cameras to perform the following tasks:

  • Path Navigation: navigate through the predefined path on the table in the meeting room.
  • Collision Avoidance: When navigating the predefined path, the Jetbot stops while detecting some specific objects in front of it and moves again while the objects are removed.
  • Object Detection in the office and the car park: to detect persons, chairs, cars and other common objects in the office and the car park.
  • Illegal Parking Detection: By using toy cars, I simulated a case that the robot could be able to detect the illegal parking case(it’s referring to that a car parks across 2 parking lots).

Hi @lzlallen1980, sounds like a very capable robot! Do you have a link to your project that you could post?