On Mon, Apr 16, 2012 at 4:40 AM, Jonathan Bohren
<jonathan.bohren@gmail.com> wrote:
On Sun, Apr 15, 2012 at 9:31 PM, Edwards, Shaun M.
<sedwards@swri.org> wrote:
I agree that hardware drivers are a good place to start. However, I think the sooner we can certify ROS core (communications, node handling, etc..), the better.
I know this is the first question I get asked by someone unfamiliar with ROS when I mention it. This is an especially important focus here at Hopkins where so many research systems are being developed for applications with high-cost-of-failure like robotic surgery, deep sea exploration, and on-orbit telerobotics.
I see two very valid and mostly orthogonal use cases so far.
1. Ensuring that certain core ROS functionalities comply with some (3rd-party certified) criteria, so that people that deal with high-cost-of-failure or safety-critical systems can trust what they are using.
2. Ensuring that a certain piece of hardware/software is ROS-compatible, so that people that venture into using them can expect compliance with specified ROS interfaces (no surprises attached).
I see great value in both, but to clarify, my initial answer to this thread was oriented mostly towards 1.
Adolfo.
-j
--
Jonathan BohrenPhD Student
Dynamical Systems and Control Laboratory
Laboratory for Computational Sensing and Robotics
The Johns Hopkins University
_______________________________________________
ros-users mailing list
ros-users@code.ros.org
https://code.ros.org/mailman/listinfo/ros-users