Ask Your Question
4

MOOS to ROS

asked 2011-07-27 00:55:19 -0600

kyle gravatar image

updated 2014-04-20 14:09:40 -0600

ngrennan gravatar image

I work for a company that deals with Autonomous Underwater Vehicles(AUV's). Currently we are in a debate over what we should use, MOOS or ROS (MOOS is just like ROS but is heavy on the underwater stuff). We would prefer to use ROS but i haven't seen many algorithms for AUV's that have already been produced in ROS. i wanted to know if there are more algorithms for AUV in ROS, other than the following: auv_joy, auv_msgs, auv_nav, auv_nav_sim,and auv_pilot.

thank you for any help.

edit retag flag offensive close merge delete

Comments

@kyle Does MOOS have a graphic interface (like gazebo or openrave) ? or is it just a control interface ?

Arkapravo gravatar imageArkapravo ( 2012-05-05 22:10:44 -0600 )edit

4 Answers

Sort by » oldest newest most voted
3

answered 2011-07-29 16:31:16 -0600

Ryan gravatar image

I attended the latest MOOS working group last week, where there was a bit of discussion on the applicability of ROS for USVs and UUVs. The biggest point that came up against using ROS for such applications is the relatively high message overhead. With acoustic modems, a lot of the assumptions used for inter-computer networking go out the window (namely, that you have >1kbps network bandwidth).

Georgia Tech is working on ROS-MOOS integration; I'm sure they've got a lot of opinions on the pros and cons: http://smartech.gatech.edu/jspui/bitstream/1853/38875/1/GT%20AUV%20Workshop%20Paper-MU3.pdf

There's also this code, though I haven't looked at it myself: http://code.google.com/p/seabee3-ros-pkg/wiki/SeaBee3

edit flag offensive delete link more

Comments

Can you say something more about the "relatively high message overhead" in ROS? Both the serialization scheme and the wire protocol(s) are intended to be simple and efficient. Is there some wasted time or space in there that we should look at?
Brian Gerkey gravatar imageBrian Gerkey ( 2011-08-02 10:43:04 -0600 )edit
1
As background information, the acoustic modem of choice for many researchers is the WHOI uModem: http://acomms.whoi.edu/umodem/, which has a speed ranging from 80 bps to 5400 bps (that's bits/sec :S). With that range, even the 38 byte/packet Ethernet frame overhead is significant.
Ryan gravatar imageRyan ( 2011-08-02 13:09:17 -0600 )edit
Now that we have rosserial, this bears re-examination, but the majority of people working in the field only know ROS in its 'typical' (read: TCPROS) form
Ryan gravatar imageRyan ( 2011-08-02 13:10:18 -0600 )edit
I have some tools for MOOS to ROS specifically: https://github.com/wjwwood/MOOS-ROS-Bridge
William gravatar imageWilliam ( 2011-10-14 17:18:12 -0600 )edit

ROS being a problem because of modem bandwidth makes no sense. Why would you send ethernet frames over a likely serial connected acoustic modem? You would write a node to convert your ROS message to modem serial input. I feel these answers are very misleading. Pls give more info

PerkinsJames gravatar imagePerkinsJames ( 2012-08-15 12:31:14 -0600 )edit

@PerkinsJames: This is exactly what rosserial is for. However, it doesn't have the flexibility in its current state.

Ryan gravatar imageRyan ( 2012-10-25 11:39:35 -0600 )edit

I agree PerkinsJames, that makes no sense. In MOOS as well, there are processes running that use acomms libraries to talk to the modem and decide what gets send across. You could do the same in ROS.

StephanieK gravatar imageStephanieK ( 2015-05-05 13:08:01 -0600 )edit
1

answered 2011-07-27 01:27:27 -0600

Stephan gravatar image

There is a simulator called "UWSim" you could check out.

UWSim visualizes an underwater virtual scenario that can be configured using standard modeling software. Controllable underwater vehicles, surface vessels and robotic manipulators, as well as simulated sensors, can be added to the scene and accessed externally through network interfaces. This allows to easily integrate the visualization tool with existing control architectures.

edit flag offensive delete link more

Comments

1
... and it has a ROS interface.
Martin Günther gravatar imageMartin Günther ( 2011-08-07 23:07:45 -0600 )edit
1

answered 2012-10-22 22:37:40 -0600

Lure_Angler gravatar image

updated 2012-10-28 17:05:00 -0600

Found this; thought it might be useful to the next guy asking this question

https://github.com/GAVLab/MOOS-ros-pkg/tree/master/MOOS


Hello Guys,

As we are also developing an AUV, we've requested for the ROS-MOOS bridge package that the GTRI folks implemented (mentioned in the paper >> "An implementation of ROS on the Yellowfin")

The link to the package is >> http://github.com/SyllogismRXS/moos-ros-bridge

Many thanks to Mr Kevin DeMarco from GTRI for kindly seeing this through.

tc

edit flag offensive delete link more
0

answered 2011-08-01 15:53:05 -0600

JonW gravatar image

I undertook a masters project to develop some AUV systems using ROS last semester. The biggest issue I had was that ROS operates in the East-North-Up frame of reference while most of the algorithms I dealt with were designed in North-East-Down.

edit flag offensive delete link more

Your Answer

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

Add Answer

Question Tools

3 followers

Stats

Asked: 2011-07-27 00:55:19 -0600

Seen: 3,404 times

Last updated: Oct 28 '12