Getting to know your Jetbot¶

Jetbot is an open source Nvidia designed educational robot https://jetbot.org/master/ . You can 3D print your own parts, print your own circuits.

Waveshare Jetbot kit¶

The Jetbot kit provided to you is one of the many third party kits available for jetbot. Waveshare documents its own kits here: https://www.waveshare.com/wiki/JetBot_AI_Kit
Waveshare’s own documentation is poor, but nvidia’s documentation is more extensive.

Features of Waveshare Jetbot kit:

• Onboard three 18650 batteries, 7800mAh, up to 12.6V voltage output, and stronger motor power.

• Onboard S-8254AA + AO4407A lithium battery protection circuit, with anti-overcharge, anti-over-discharge, anti-over-current and short-circuit protection functions.

• The onboard APW7313 voltage regulator chip can provide a stable 5V voltage to the Jetson Nano.

• The onboard TB6612FNG dual H-bridge motor driver chip can drive the left and right motors to work.

• Onboard 0.91-inch 128 × 32 resolution OLED, real-time display of car IP address, memory, power, etc.

• The onboard ADS1115 AD acquisition chip is convenient for real-time monitoring of battery power.

On board computer¶

We are using Jetson Nano: P3450 developer kit as the onboard computer for the jetbot. The documentation is here.

Preinstalled software: Jetpack SDK 4.5.1¶

Nvidia releases supported software for Jetson Nano developer kits as Jetpack SDK and Jetpack Linux. We have Jetpack SDK version 4.5 whose documentation is available here.

1. OS > Linux > Ubuntu 18.04: We are running Nvidia’s Linux for Tegra (L4T) R32.51 that is the “board support package” for Jetson Nano. It includes Linux Kernel 4.9, bootloader, NVIDIA drivers, flashing utilities, sample filesystem based on Ubuntu 18.04 (bionic), and more for the Jetson platform. Full documentation is here: https://docs.nvidia.com/jetson/archives/l4t-archived/l4t-3251/index.html

2. JetPack components

   1. Deep learning libraries:

      1. TensorRT: A version of Tensorflow for embedded devices

      2. CUDA: Programming library for Nvidia GPU

      3. cuDNN: Libraries for Deep Neural Network components with CUDA

   2. Multimedia API: Libraries for video and image processing

   3. VisionWorks, OpenCV, VPI: Implementations of computer vision algorithms.

3. Docker
   Docker is similar to a virtual machine engine, but it is more closely integrated with the host operating system, hence much lighter than a virtual machine. Jetpack comes with a docker installation and with two docker images that run on startup.

   Docker images: You can view the docker containers running on your jetbot by typing

jetbot@nano-4gb-jp45:~$ sudo docker container ls
CONTAINER ID   IMAGE                                COMMAND
CREATED     STATUS      PORTS   NAMES
9e3701e9ef68   jetbot/jetbot:jupyter-0.4.3-32.5.0   "/bin/sh -c 'cd $JUP…"   2
years ago   Up 7 minutes            jetbot_jupyter
198acf2fea98   jetbot/jetbot:display-0.4.3-32.5.0   "/bin/sh -c 'python3…"   2
years ago   Up 7 minutes            jetbot_display

There are two docker containers running jetbot_jupyter and jetbot_display. The first one runs the jupyter notebooks that you can access through `http://<jetbot_ip_address>:8888`. The second one runs the code necessary to run OLED display stats. You can modify this code in `/home/jetbot/jetbot/jetbot/apps/stats.py`

You can learn more about docker and its commands here: https://docs.docker.com/get-started/