NVIDIA-Jetson / redtail - Wiki

NVIDIA-Jetson / redtail - Wiki

https://github.com/NVIDIA-Jetson/redtail/wiki

Perception and AI components for autonomous mobile robotics.

Home

Introduction

This project contains instructions, code and other artifacts that allow users to build mobile robots (drones, rovers) which can autonomously navigate through highly unstructured environments like forest trails. Our components use deep learning-based AI running on an NVIDIA Jetson embedded platform. The code implements ideas discussed in the arXiv paper, see references section.

The project has two major parts, modeling and platform implementation.

Modeling

The project’s AI that enables autonomous navigation is based on a deep neural network (DNN) which can be trained from scratch using publicly available data. A few pre-trained DNNs are also available as a part of this project. In case you want to train TrailNet DNN from scratch, follow the steps on this page.
https://github.com/NVIDIA-Jetson/redtail/tree/master/models/pretrained
https://github.com/NVIDIA-Jetson/redtail/wiki/Models

Platforms

The following platforms are currently supported:

TBS Discovery Platform
3DR IRIS+, older
Differential drive rover support is coming
https://github.com/NVIDIA-Jetson/redtail/wiki/Skypad-TBS-Discovery-Setup
https://github.com/NVIDIA-Jetson/redtail/wiki/3DR-Iris-Setup

In general, any platform that uses the Pixhawk autopilot should work too.
https://pixhawk.org/

There is also an experimental support for APM Rover:

Erle Rover
https://github.com/NVIDIA-Jetson/redtail/wiki/Erle-Rover-Setup

Getting started

入门指南

Building a complete autonomous drone platform requires proper hardware and software configuration.

Jetson setup

The NVIDIA Jetson platform is used to run most of the components, such as DNN inference, the controller, and video streaming. The Jetson setup guide describes steps to install all required software and dependencies.
https://github.com/NVIDIA-Jetson/redtail/wiki/Jetson-Setup

Drone setup

Depending on the drone platform some additional steps might be required. Follow the steps in the documentation for your particular platform:

TBS Discovery Platform
3DR IRIS+, older
https://github.com/NVIDIA-Jetson/redtail/wiki/Skypad-TBS-Discovery-Setup
https://github.com/NVIDIA-Jetson/redtail/wiki/3DR-Iris-Setup

GCS (Ground Control Station) setup

A laptop is a convenient way to run GCS software like QGroundControl as well to control the drone using an NVIDIA Shield or an XBox controller. Follow these steps to setup GCS machine.

Simulation

It is usually a good idea to test your code in a simulator. Follow these steps to run simulations using Gazebo.
http://qgroundcontrol.com/
https://github.com/NVIDIA-Jetson/redtail/wiki/GCSSetup

Flying

Once the hardware and software setup steps are complete, it’s time to take off! Follow these steps to fly the drone.
https://github.com/NVIDIA-Jetson/redtail/wiki/Launch-Sequence-and-Flying

References

[Toward Low-Flying Autonomous MAV Trail Navigation using Deep Neural Networks for Environmental Awareness]

Our GTC 2017 talk: slides, video
http://on-demand.gputechconf.com/gtc/2017/presentation/s7172-nikolai-smolyanskiy-autonomous-drone-navigation-with-deep-learning.pdf
http://on-demand.gputechconf.com/gtc/2017/video/s7172-smolyanskiy-autonomous-drone-navigation-with-deep-learning%20(1).PNG.mp4

Demo video showing 250 m autonomous flight, DNN activation and control
https://www.youtube.com/watch?v=H7Ym3DMSGms

Demo video showing our record making 1 kilometer autonomous flight
https://www.youtube.com/watch?v=USYlt9t0lZY

Demo video showing generalization to ground vehicle control and other environments
https://www.youtube.com/watch?v=ZKF5N8xUxfw

Stereo DNN, GTC18 talk: arXiv paper, Stereo DNN video demo
[On the Importance of Stereo for Accurate Depth Estimation: An Efficient Semi-Supervised Deep Neural Network Approach]
https://www.youtube.com/watch?v=0FPQdVOYoAU&feature=youtu.be

Wordbook

autonomous [ɔː’tɒnəməs]：adj. 自治的，自主的，自发的
robotic [ro’bɑtɪk]：adj. 机器人的，像机器人的，自动的 n. 机器人学
drone [drəʊn]：n. 雄蜂，嗡嗡的声音，懒惰者 n. 无人机 (非正式) vi. 嗡嗡作声，混日子 vt. 低沉地说
artifact ['ɑ:təˌfækt]：n. 人工制品，手工艺品
forest trail：林荫小径，林间小径
sidewalk ['saɪdwɔːk]：n. 人行道
ground control station，GCS：地面控制站
GPU Technology Conference，GTC：GPU 技术大会
rover ['rəʊvə]：n. 漫游者，流浪者，漂泊者
unmanned ground vehicle，UGV：无人地面车辆
Team BlackSheep，TBS：黑羊队