Ask Your Question

what is the safety function for ROS communication?

asked 2020-08-19 04:07:57 -0500

castpuz gravatar image

I'm looking for the safety function like diagnostics system or message_filter.

I would like to know about other function that ensures the safety for ROS based robot system.

Is there a function like this?

edit retag flag offensive close merge delete


I dont think ROS has much to offer, but ROS2 introduced the "managed node" concept link text , but maybe I took the question the wrong way?

Dragonslayer gravatar image Dragonslayer  ( 2020-08-19 06:54:25 -0500 )edit

Dragonslayer Thank you for your reply. It is certainly a function of the stability of the system. Thank you. Do you know any other features?

castpuz gravatar image castpuz  ( 2020-08-24 09:27:24 -0500 )edit

You mentioned the watchdog timer in the other comment already, besides that I only red about people speculating about building their own system management stuff, but it likely introduces more instability than it solves. (specially if want to make it comply with ISOs and verification and stuff) ROS2 seems the most easy way to go, it was made with better stability in mind. But as fergs has already written you might sideline the whole issue with a hardware safety controller. I dont see ROS park my car relying on itself.

Dragonslayer gravatar image Dragonslayer  ( 2020-08-25 10:22:46 -0500 )edit

Dragonslayer Thank you. In short, ROS1 doesn't have an official tool for safety functions? I'm sad to hear that.

castpuz gravatar image castpuz  ( 2020-09-09 23:40:01 -0500 )edit

@castpuz: ROS1 was never designed for the use-cases you seem to have in mind.

It is a platform primarily created for research purposes, which has certainly been used by companies to create products and very successfully so, but the design never had productisation, commercialisation or certification in mind.

ROS 2 -- sometimes called "ROS for products" -- is entirely different.

You may want to look into that.

gvdhoorn gravatar image gvdhoorn  ( 2020-09-09 23:44:22 -0500 )edit

1 Answer

Sort by ยป oldest newest most voted

answered 2020-08-22 19:51:49 -0500

fergs gravatar image

ROS 1 really doesn't have any built-in "functional safety". Typically, this is handled below the level of ROS - for instance, a mobile robot which complies with ISO 13849 will typically use a safety laser scanner with a hardware safety controller (I've previously used a SICK S300 and FlexiSoft) and then have ROS do the high level navigation. As long as the high level navigation is working OK, you never "hit" the safety limits being enforced by the lower level hardware. Should something die in ROS, the functionally safe components below do their job as designed and certified. This also means that the ROS-level code can change more often as it is NOT part of the safety assessment.

edit flag offensive delete link more


Thank you for your reply. I know that even within ROS1, functions such as watchdog timer have been implemented for the stability of the system. Do you happen to know the functions implemented in a similar ROS?

castpuz gravatar image castpuz  ( 2020-08-24 09:26:08 -0500 )edit

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

Question Tools



Asked: 2020-08-19 04:07:57 -0500

Seen: 125 times

Last updated: Aug 22 '20