Autoware on Yonohub — Part 1
We are pleased to announce the release of AutowareAI ready-to-use environment on Yonohub. With this environment, you can use Autoware.AI with all of its modules directly on Yonohub without downloading it or setting up any special software. Seamlessly integrate your algorithms and connect them to any dataset or a ROSBag. Sign up now and start using Autoware.AI ready-to-use environment.
Learning Objectives
- What is Autoware?
- Why use Autoware on Yonohub?
- Creating AutowareAI Custom App
Introduction
Autoware was founded in 2015 by Shinpei Kato (Still in the Autoware Foundation board) from Nagoya University in Japan. It was designed as an “All-in-One” open-source software for autonomous driving technology. In 2018, the Autoware Foundation was founded as a non-profit organization to support open-source projects to support the development and implementation of autonomous driving solutions.
Actually, Autoware has two stacks, one is AutowareAI based on ROS (will be sundown end of 2020), and AutowareAuto is the latest Stack based on ROS2.
AutowareAI was the first Autoware project based on ROS1, and the Apache License 2.0 includes Localization, Detection, Prediction, and Planning.
AutowareAuto is the latest Stack based on ROS2, which should soon integrate all Features of AutowareAI, plus clear defined APIs and Interfaces to be more modular and flexible to be able to integrate more Sensors. In addition, it offers the latest Software Development Technologies from continuous development and integration.
Autoware already caught a lot the attention of autonomous car researchers and companies around the world and is also used by more than 100 companies and run on more than 30 vehicles. Apex.AI uses AutowareAuto in Apex.os, which was founded by ex Bosch Employee Jan Becker.
Autoware is supported by a large community of developers and Universities like the University in Aachen and Graz, for example.
Check out this video from runs with AutowareAuto:
This Video gives you a great impression of how seamless the vehicle (Toyota Lexus equipped with AutowareAuto) is successfully moving through a city managing all driving scenarios by itself.
The installation for Autoware on a local machine with all its dependencies like Docker and ADE could take a long time and requires high hardware specifications and an NVIDIA graphics card min GTX 1080. You want to overcome these requirements, here is the solution Autoware on Yonohub Cloud Environment. www.yonohub.com.
The advantage here is:
- no own hardware necessary (CPUs, GPUs, SSD, cc..).
- use ready environments and frameworks or design your own.
- pay only when you use it.
In this article, we show an example of Autoware.AI usage on Yonohub, and in the upcoming articles, we’ll discuss other usages of Autoware.AI modules.
AutowareAI Custom App
We’ve made a ready-to-use environment for AutowareAI and you can find it on YonoStore. but if you are looking forward to building your own AutowareAI environment to use on a custom app, the following steps will show you how to do that.
Step 1: Create a new environment
Using YonoEBuilder, you can create a new environment based on the Nvidia Docker image (nvidia/cuda:10.0-cudnn7-devel-ubuntu18.04) and then click on (Custom App Support — Visual Studio Code — NoVNC) then click on “Express Build”.
Step 2: Open AutowareAI environment
After building the new environment from the previous step, you’ll need to open the environment as a custom app on Yonohub. So, in order to do that, you’ll click on the “+” sign in the main view and create a custom app as follows.
Step 3: Build AutowareAI
After Launching the AutowareAI custom app, you can use either visual studio code or NoVNC to build Autoware.ai from source by following this link and also you’ll need to install other libraries if you want to use all the modules inside Autoware.ai such as TensorRT, SSDCaffe, and ENet. You can also find info about the installation of those libraries in this link.
Step 4: Save AutowareAI Environment
After Installing AutowareAI, you can save the new environment as a new version by doing the following steps:
Also, you can make other changes after that and save other versions of the environment if you want.
In the next article, I’ll show you how to use Autoware.ai algorithms on YonoArc as blocks to make it easier and faster to develop your pipelines.
Yonohub
Yonohub is the first cloud-based system for designing, sharing, and evaluating autonomous vehicle algorithms using just blocks. Yonohub features a drag-and-drop tool to build complex systems consisting of many blocks, a marketplace to share and monetize blocks, a builder for custom environments, and much more.
Get $25 free credits when you sign up now. For researchers and labs, contact us to learn more about Yonohub sponsorship options. Yonohub: A Cloud Collaboration Platform for Autonomous Vehicles, Robotics, and AI Development. www.yonohub.com
If you liked this article, please consider following us on Twitter at @yonohub, email us directly, or find us on LinkedIn. I’d love to hear from you if I can help you or your team with how to use YonoHub.
Special thanks to Alexander Kolbai for his help and support in this article.
References
https://www.autoware.ai/
https://gitlab.com/autowarefoundation/autoware.ai
