| ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange |
Which would be having more support? Can anyone point to useful resources for both of them? For the Autoware.Auto I found the ApexAI courses. Our team is more well versed with ROS1 but if Autoware.AI won't have that much support we can make the switch. We will mainly be building a robot for the LGSVL simulator and navigating through the San Francisco environment
I m using both of them.
From my experience AutowareAuto offers more tutorials and courses. But the flexibility to write codes, reuse packages, switch sensor, planners, etc are not possible without hard implementation from your team side. Besides that AutowareAuto is 90% C++ based. While in AutowareAI you have possibility to work more with Python, you have also more flexibility and an outstanding user-interface (The runtime-manager) that AutowareAuto does not provide. Therefore your team can control the modules easier and have more flexibility to switch the algorithms of each one of these modules (perception, localization, routing, etc).
I would say AutowareAI is best for survey purposes. AutowareAuto howeever is more suitable for application and test on the real autonomous vehicle, once it assures more safety on connections.
Autoware.AI has entered its maintenance phase, like ROS1: https://discourse.ros.org/t/end-of-li...
So it's safe to say that Autoware.Auto + ROS2 will be better-supported in the future. The main docs for Autoware.Auto are at https://autowarefoundation.gitlab.io/..., although they're currently being updated.
Asked: 2020-12-06 22:14:37 -0500
Seen: 2,078 times
Last updated: Apr 09 '21
ROS Answers is licensed under Creative Commons Attribution 3.0 Content on this site is licensed under a Creative Commons Attribution Share Alike 3.0 license.
