Jetson AI Fundamentals - S2E1 - JetBot Intro and Hardware

NVIDIA Developer · Beginner ·📐 ML Fundamentals ·5y ago

Key Takeaways

The Jetson AI Fundamentals series introduces the JetBot, a deep learning machine that can be built using the Jetson Nano and various hardware components, with tutorials on hardware assembly and software setup using tools like Docker containers and Adafruit JetBot kits.

Full Transcript

hello it's jim from jetsonhacks.com this is the start of a jetson ai fundamental series on the jetbot an easy to build deep learning machine [Music] the jetbot series will consist of several episodes in the first episode this episode we will build the hardware it will look very similar if not exactly like this when we are done in the second episode we will install the software on the robot that will include a docker container which contains all of the deep learning libraries that we will need for the robot to learn its environment in later episodes we will learn how to use deep learning to solve some common robotics problems such as collision avoidance and road following let's here's a green jetbot it's one of the original jetbots that they built at nvidia on the top of the robot we have our jetson nano it is one of the original jetson a02 models at the front of the robot we have a camera it has 145 degrees field of view it's angled slightly downward so that it can see in front of itself a little bit better connected to the gpio pins is an oled display it shows information about the jetbot including the wi-fi address from there the signals from the gpio header are extended so that we can talk to the motors power is supplied to the jetson via the micro usb port there are two usb plugs on the battery one powers the jetson and the other powers the motors let's turn the jetbot over underneath the jetbot you can see the two motors they are connected to the wheels and here's the motor driver the motor driver consists of an i2c to pwm driver which talks to two h bridges the h bridges control the motors the jets and nano speaks i2c the converter converts it to pwm so that the h-bridge drivers can understand it the jetbot only uses one of the h-bridges and then here on the back is a little caster ball it's very slippery it helps the jetbot to balance the jet bot is a differential drive robot it steers by moving the wheels at different speeds the jetbot can go forwards and backwards the jetbot wiki is where we go for all things chatbot go here to get the latest build instructions and all the up-to-date informations if you are building your own a jet pot here's the bill of materials this lists all the parts and some buying options there are also instructions for building the hardware here are some tools you will need here are all the steps to actually build the robot it's very comprehensive along with pictures i like pictures you should be able to follow these and have a working jetbot at the end here's how you mount the motor driver for example wi-fi antennas it tells you how to install your wi-fi card latest and greatest instructions this is where you go anyway go to this page and follow along when you build the project is open source and there's a pretty good size community to support it there are also sections for software setup and there are examples of course we will be covering some of those in the upcoming episodes there are also third-party kits available which contain all of the parts to build a jet bot definitely worth checking out let's take a quick run through some of them here's wave share you can get this through amazon even though these look a little bit different than the jetbot that we will be building they all run the same software sparkfun has a version 2.1 available seed studio silicon highway has them they are a european distributor for the jetsons here's the fabo version this one's kind of fun it's yellow red and blue it looks like a traditional jetbot the donkey car store has the chassis available if you need to find one the jetbot wiki is the place to start it's also where the code is kept that's important too let's start our build we did the traditional way we printed out our own chassis the 3d print files are available on the jetbot wiki there are four parts this is the camera holder these two parts hold the caster ball and this is the main chassis here's the best part let's pull it off the built plate they came out pretty nice this came out as a pretty good print printed out in high quality so let's finish cleaning this up and we can start installing components for this build we're using the adafruit jetbot kit includes all of these parts two motors two wheels an led display some headers and a motor driver we will follow the adafruit instructions and solder the pins to the driver board now let's prepare our motors to prepare the motor remove the original zip tie then remove the clear plastic retainer it needs a little bit of persuasion to remove it rearrange the red and black wires so that you can protect them with the piece of plastic when you reinstall it and then you can attach the motors with 25 millimeter screws it's pretty simple when dealing with screws and small parts use a compartmentalized placemat or a little container that keeps the screws off the floor your vacuum cleaner will thank you prepare the driver board according to the adafruit instructions strip the usb cable according to the instructions on the jetbot wiki this requires a little bit of soldering if you've never taken apart a usb cable before there's a little bit of shielding around the wires that you'll need to cut through and remove these are pretty small wires so you have to be a little bit careful strip a little bit off the end and then insert it into the screw terminal i used a little bit of heat shrink wrapping here to help insulate things make sure that you run the cable through the chassis before you mount the driver and then we use these self-tapping screws to attach the driver to the chassis the motor driver has been secured and i wired the motor the pictures on the wiki are wrong make sure that you read the directions it should be red black red black and now we install the caster ball it's very slippery it's kind of fun it's made of delrin feels like a shell game or something make sure that the ball rolls freely and then secure it with the self-tapping screws next we mount the tires on the wheels then we will mount the wheels on the robot the instructions do not say to do this but i think it would be a valuable addition to our robot installation complete the wheels are now on let's take a look at them it's starting to look like a robot we are ready to install our jetson we are using something a little bit special today we are using one of the new jets and nano 2gbs we have the wi-fi dongle installed and then we attach it to the jetbot using four self tapping screws with the jetson attached to the chassis now we can install the pi oled display i soldered on the right angle header to the oled one thing you may find useful is to use one of the double-sided tapes from the kit and attach the display to something to give you a stable base to solder from this is a little bit challenging use a small tip on your soldering iron try to approach it from the back so you don't touch the oled follow the adafruit instructions that way you don't have to worry about ruining the display next attach the camera to the camera mount and then the camera mount to the chassis using eight self tapping screws we're nearing the home stretch our camera is now installed we will have to remember to remove the lens cap i found that you get a better image that way we are now ready to install the battery we use a couple pieces of double-sided tape to hold the battery in place okay our batteries in place we plugged in our motors not quite sure how this happened but somehow googly eyes got on here we have one more cable to install and that goes from the usb battery port to the usb c connector on the jetson nano 2 gb and then we are ready to run overall this is a pretty easy starter build this is kind of a beginner maker level type of project gives you a good chance to practice your soldering skills but there's not a whole lot of it so it's not tiring and most of the mechanical bits and pieces are pretty forgiving in our next episode in the jetson ai fundamentals jetbot series we will be installing the software and making the motors run on the jetbot hey thanks for watching to learn more visit nvidia.com dli or email us at nvdli at nvidia.com

Original Description

The Jetson Nano JetBot is a great introduction to robotics and deep learning. There are several options in building your own hardware, here we share some valuable tips building yours. 00:00:00 - Introduction - JetBot course 00:00:58 - Introduction - JetBot hardware 00:02:48 - Introduction - JetBot Wiki 00:05:00 - Third party JetBot kits 00:06:31 - Building a JetBot from scratch 00:07:29 - Installing components to the JetBot chassis 00:12:24 - Conclusion JetBot - https://github.com/NVIDIA-AI-IOT/jetbot Jetson AI Fundamentals - https://nvidia.com/jetson-ai-fundamentals NVIDIA Developer Forums - https://forums.developer.nvidia.com
Watch on YouTube ↗ (saves to browser)
Sign in to unlock AI tutor explanation · ⚡30

Playlist

Uploads from NVIDIA Developer · NVIDIA Developer · 57 of 60

1 Ray Tracing Essentials Part 2: Rasterization versus Ray Tracing
Ray Tracing Essentials Part 2: Rasterization versus Ray Tracing
NVIDIA Developer
2 Ray Tracing Essentials Part 3: Ray Tracing Hardware
Ray Tracing Essentials Part 3: Ray Tracing Hardware
NVIDIA Developer
3 Ray Tracing Essentials Part 4: The Ray Tracing Pipeline
Ray Tracing Essentials Part 4: The Ray Tracing Pipeline
NVIDIA Developer
4 NsightGraphics 2020 2 Release Spotlight
NsightGraphics 2020 2 Release Spotlight
NVIDIA Developer
5 Ray Tracing Essentials Part 5: Ray Tracing Effects
Ray Tracing Essentials Part 5: Ray Tracing Effects
NVIDIA Developer
6 Ray Tracing Essentials Part 6: The Rendering Equation
Ray Tracing Essentials Part 6: The Rendering Equation
NVIDIA Developer
7 Ray Tracing Essentials Part 7: Denoising for Ray Tracing
Ray Tracing Essentials Part 7: Denoising for Ray Tracing
NVIDIA Developer
8 Spatiotemporal Importance Resampling for Many-Light Ray Tracing (ReSTIR)
Spatiotemporal Importance Resampling for Many-Light Ray Tracing (ReSTIR)
NVIDIA Developer
9 Announcing Cloud-Native Support for Jetson Platform
Announcing Cloud-Native Support for Jetson Platform
NVIDIA Developer
10 JetsonTV: Build your next project with NVIDIA Jetson
JetsonTV: Build your next project with NVIDIA Jetson
NVIDIA Developer
11 Nsight Compute Feature Spotlight: Roofline Analysis, Asynchronous Copy, Sparse Data Compression
Nsight Compute Feature Spotlight: Roofline Analysis, Asynchronous Copy, Sparse Data Compression
NVIDIA Developer
12 Nsight Systems Feature Spotlight: OpenMP
Nsight Systems Feature Spotlight: OpenMP
NVIDIA Developer
13 Isaac Sim 2020: Deep Dive
Isaac Sim 2020: Deep Dive
NVIDIA Developer
14 NVIDIA Jetson: Enabling AI-Powered Autonomous Machines at Scale
NVIDIA Jetson: Enabling AI-Powered Autonomous Machines at Scale
NVIDIA Developer
15 NVIDIA Tools to Train, Build, and Deploy Intelligent Vision Applications at the Edge
NVIDIA Tools to Train, Build, and Deploy Intelligent Vision Applications at the Edge
NVIDIA Developer
16 Jetson Xavier NX Developer Kit: The Next Leap in Edge Computing
Jetson Xavier NX Developer Kit: The Next Leap in Edge Computing
NVIDIA Developer
17 Synthesizing High-Resolution Images with StyleGAN2
Synthesizing High-Resolution Images with StyleGAN2
NVIDIA Developer
18 NVIDIA Robotics: Isaac SDK and Sim 2020.1
NVIDIA Robotics: Isaac SDK and Sim 2020.1
NVIDIA Developer
19 Accelerating COVID-19 Research with GPUs
Accelerating COVID-19 Research with GPUs
NVIDIA Developer
20 Visualizing 150 Terabytes of Data
Visualizing 150 Terabytes of Data
NVIDIA Developer
21 Boosting Performance and Utilization with Multi-Instance GPU
Boosting Performance and Utilization with Multi-Instance GPU
NVIDIA Developer
22 Running Multiple Workloads on a Single A100 GPU
Running Multiple Workloads on a Single A100 GPU
NVIDIA Developer
23 NVIDIA Nsight Feature Spotlight: GPU Trace
NVIDIA Nsight Feature Spotlight: GPU Trace
NVIDIA Developer
24 Spark 3 Demo: Comparing Performance of GPUs vs. CPUs
Spark 3 Demo: Comparing Performance of GPUs vs. CPUs
NVIDIA Developer
25 NVIDIA Jetson Nano Wins Edge AI and Vision Alliance Award
NVIDIA Jetson Nano Wins Edge AI and Vision Alliance Award
NVIDIA Developer
26 NVIDIA IndeX on Google Cloud Platform Marketplace
NVIDIA IndeX on Google Cloud Platform Marketplace
NVIDIA Developer
27 DeepStream SDK: Best practices for performance optimization
DeepStream SDK: Best practices for performance optimization
NVIDIA Developer
28 Efficiently Deploying GPU Accelerated 5G CloudRAN for Edge AI Inferencing
Efficiently Deploying GPU Accelerated 5G CloudRAN for Edge AI Inferencing
NVIDIA Developer
29 NVIDIA PhysicsNeMo - Accelerating Scientific & Engineering Simulation Workflows with AI
NVIDIA PhysicsNeMo - Accelerating Scientific & Engineering Simulation Workflows with AI
NVIDIA Developer
30 NVIDIA Deep Learning Institute Instructor-Led Training Available Remotely
NVIDIA Deep Learning Institute Instructor-Led Training Available Remotely
NVIDIA Developer
31 Advancing AR Glasses
Advancing AR Glasses
NVIDIA Developer
32 Blender Cycles: RTX On
Blender Cycles: RTX On
NVIDIA Developer
33 Real-Time GPU-Accelerated Data Analytics of 250 million Flight Data Records of 737 Max grounding
Real-Time GPU-Accelerated Data Analytics of 250 million Flight Data Records of 737 Max grounding
NVIDIA Developer
34 Assessing Property Damage with AI
Assessing Property Damage with AI
NVIDIA Developer
35 RAPIDS: GPU-Accelerated Data Analytics & Machine Learning
RAPIDS: GPU-Accelerated Data Analytics & Machine Learning
NVIDIA Developer
36 DaVinci Resolve Turns RTX On
DaVinci Resolve Turns RTX On
NVIDIA Developer
37 RAPIDS with Plotly Dash : GPU-Accelerated Census 2010 Visualization
RAPIDS with Plotly Dash : GPU-Accelerated Census 2010 Visualization
NVIDIA Developer
38 NVIDIA IndeX for arivis5D Cloud Platform
NVIDIA IndeX for arivis5D Cloud Platform
NVIDIA Developer
39 NVIDIA Backchannel: Behind the Scenes of Marbles at Night RTX
NVIDIA Backchannel: Behind the Scenes of Marbles at Night RTX
NVIDIA Developer
40 NVIDIA Backchannel: Sneak Peek into Marbles RTX in Omniverse
NVIDIA Backchannel: Sneak Peek into Marbles RTX in Omniverse
NVIDIA Developer
41 How to Create "Paint" in Substance Painter
How to Create "Paint" in Substance Painter
NVIDIA Developer
42 Accelerate AI development for Computer Vision on the NVIDIA Jetson with alwaysAI
Accelerate AI development for Computer Vision on the NVIDIA Jetson with alwaysAI
NVIDIA Developer
43 Securing Next Generation Apps over VMware Cloud Foundation with Bluefield-2 DPU
Securing Next Generation Apps over VMware Cloud Foundation with Bluefield-2 DPU
NVIDIA Developer
44 Accelerated Data Centers with NVIDIA and VMware
Accelerated Data Centers with NVIDIA and VMware
NVIDIA Developer
45 GPU-Accelerated Motion Blur in Blender Cycles
GPU-Accelerated Motion Blur in Blender Cycles
NVIDIA Developer
46 NVIDIA Clara Guardian Virtual Patient Assistant
NVIDIA Clara Guardian Virtual Patient Assistant
NVIDIA Developer
47 Revolutionizing Supercomputing with NVIDIA UFM Cyber-AI
Revolutionizing Supercomputing with NVIDIA UFM Cyber-AI
NVIDIA Developer
48 Inventing Virtual Meetings of Tomorrow with NVIDIA AI Research
Inventing Virtual Meetings of Tomorrow with NVIDIA AI Research
NVIDIA Developer
49 Learning a Contact-Adaptive Controller for Robust, Efficient Legged Locomotion
Learning a Contact-Adaptive Controller for Robust, Efficient Legged Locomotion
NVIDIA Developer
50 Getting started with Jetson Nano 2GB Developer Kit
Getting started with Jetson Nano 2GB Developer Kit
NVIDIA Developer
51 NVIDIA Jetson Developer Community AI Projects
NVIDIA Jetson Developer Community AI Projects
NVIDIA Developer
52 Open-source projects on NVIDIA Jetson Nano 2GB Developer Kit
Open-source projects on NVIDIA Jetson Nano 2GB Developer Kit
NVIDIA Developer
53 Real-Time Ray Tracing with Project Lavina
Real-Time Ray Tracing with Project Lavina
NVIDIA Developer
54 Jetson AI Fundamentals - S1E2 - Hello Camera
Jetson AI Fundamentals - S1E2 - Hello Camera
NVIDIA Developer
55 Develop Optimized Conversational AI Models with NVIDIA NeMo on DGX A100
Develop Optimized Conversational AI Models with NVIDIA NeMo on DGX A100
NVIDIA Developer
56 Jetson AI Fundamentals - S1E4 - Image Regression Project
Jetson AI Fundamentals - S1E4 - Image Regression Project
NVIDIA Developer
Jetson AI Fundamentals - S2E1 - JetBot Intro and Hardware
Jetson AI Fundamentals - S2E1 - JetBot Intro and Hardware
NVIDIA Developer
58 Jetson AI Fundamentals - S2E2 - JetBot Software Setup
Jetson AI Fundamentals - S2E2 - JetBot Software Setup
NVIDIA Developer
59 Jetson AI Fundamentals - S1E1 - First Time Setup with JetPack
Jetson AI Fundamentals - S1E1 - First Time Setup with JetPack
NVIDIA Developer
60 Jetson AI Fundamentals - S1E3 - Image Classification Project
Jetson AI Fundamentals - S1E3 - Image Classification Project
NVIDIA Developer

This video introduces the JetBot, a deep learning machine that can be built using the Jetson Nano and various hardware components, and provides tutorials on hardware assembly and software setup. The JetBot is a great introduction to robotics and deep learning, and can be used to solve common robotics problems like collision avoidance and road following. By following the steps in this video, viewers can build their own JetBot and start exploring the world of deep learning and robotics.

Key Takeaways
  1. Build the hardware
  2. Install the software on the robot
  3. Use deep learning to solve common robotics problems
  4. Talk to the motors
  5. Power the Jetson and the motors
  6. Prepare the motor by removing the original zip tie and clear plastic retainer
  7. Rearrange the red and black wires
  8. Attach the motors with 25 millimeter screws
  9. Prepare the driver board according to the Adafruit instructions
  10. Strip the USB cable according to the instructions on the JetBot wiki
💡 The JetBot is a great introduction to robotics and deep learning, and can be used to solve common robotics problems like collision avoidance and road following using deep learning libraries in a Docker container.

Related AI Lessons

10 Python Concepts You Must Know Before Calling Yourself Advanced
Learn 10 essential Python concepts to take your skills to the advanced level and stand out as a developer
Medium · AI
10 Python Concepts You Must Know Before Calling Yourself Advanced
Learn 10 crucial Python concepts to elevate your skills from intermediate to advanced and become a proficient developer
Medium · Data Science
10 Python Concepts You Must Know Before Calling Yourself Advanced
Learn 10 essential Python concepts to take your skills to the advanced level and stand out as a developer
Medium · Programming
10 Python Concepts You Must Know Before Calling Yourself Advanced
Learn 10 essential Python concepts to take your skills to the advanced level and separate yourself from beginner developers
Medium · Python

Chapters (7)

Introduction - JetBot course
0:58 Introduction - JetBot hardware
2:48 Introduction - JetBot Wiki
5:00 Third party JetBot kits
6:31 Building a JetBot from scratch
7:29 Installing components to the JetBot chassis
12:24 Conclusion
Up next
Is Python Dead in 2026?| Truth About Python in AI Era | 90 Days Roadmap @FameWorldEducationalHub
FAME WORLD EDUCATIONAL HUB
Watch →