[Fiware-robotics] R: R: Live Demo phase 1

[Antonini Roberto]

My answer online ...

Da: fiware-robotics-bounces at lists.fi-ware.org [mailto:fiware-robotics-bounces at lists.fi-ware.org] Per conto di Bollano Gianmario
Inviato: giovedì 16 aprile 2015 16:40
A: fiware-robotics at lists.fi-ware.org
Oggetto: [Fiware-robotics] R: Live Demo phase 1

Hi all,
thank you Davide for the contributions (docs and slide) and Inigo for the explanation about the tests done with Kurento.

Below are some interesting points arising from this discussion and the last Robotics call:


1.       We have decided to use Google Docs to share information and  speedup development. For what concern the FIROS-RCM communication.

a.       If you you agree I would suggest to use and update the proposal in the google doc FIROS_RCM_v3.0.docx<https://docs.google.com/document/d/1eWO5MYYkjJEo1xLLm2YsX_bhCnET5eX4dm4_cL9MFg0/edit?pli=1> prepared by Davide.

b.      We still have to define more in detail the format of the messages we would use in the first demo, specifically messages to control the robot movement:

·   how many types of movement?
R: For the first phase, I would only consider differential drive robots

·    which type of control needed? (it may depends on Robots type)
R: ROS abstraction typically suggests geometry_msgs/Twist ROS type message to control robot movement as a standard de-facto, if different ROS message types were requested, an adaption is needed which reflects that message to CB entity attribute making use of Twist type, ET correct me if I'm wrong.

·   Will we have controls also for the movement of the camera?
R: For the very first phase, I wouldn't consider a PTZ camera, if desired then a ROS (or to be ROSified) PTZ command is needed, the last part of above consideration/answer could be applied also for this answer: you (ET) need to take into consideration a new message type to control PTZ camera.


c.       We may need further messages, beside those already defined till now:

·   RCM_Event for connection/disconnection of robots,

·   FIROS_Info for message type and communication type supported by any Robot,

·   CB_Event when instances are created or removed from CB
R: in my opinion, these message should  suffice our needs, your (of ET) considerations are warmly welcomed.
For instance, a message might be required after RCM master launch FIROS, when FIROS is up and working.

d.      If we want to control more than one robot, let's say at least 2 robots (as in the requirements of the demo) , we still have to decide if we have

·   one FIROS instance for all robots and how name unambiguously the robot names,

·   or more FIROS instances (with different names), one for robot or maybe one for location if robots are in different sites



2.       Kurento: Davide and Inigo have tried two different setup. We are free to define the best setup for the demo  but there are some point we need to consider carefully:

a.       One of the WebRTC endpoint (for video transmission) is in the Robot, thus the simplest solution might be chrome in a Linux environment. Having KAS (with node.js) is maybe possible but it might require excessive resources for a Linux environment hosted on a robot.

b.      For the other WebRTC endpoint (where the Control demo GUI app resides) we do not need chrome but obviously we should have stream compatibility with the other Chrome (on the Robot) endpoint. That might imply format and throughput of video stream.

c.       The KAS (Kurento App Server) might be hosted in the same machine with the RCM_Master. This machine might need to receive commands from the Control Demo Gui App then it should have public  address if we want the Control GUI to register dynamically as webRTC endpoint.

d.      The KMS ideally should be the one in FIWARE lab, because that's one of the keypoint of the live demo:give example of usage of  GEs available be the FIWARE lab. If however we'll find trouble in using the kurento filters, we can stick to a local KMS.

e.      Do we need Kurento filters for the first Teleoperation demo?

f.        If we don't need any specific kurento filter, shouldn't we use a Kurento Video-transmission widget made available by Wirecloud instead? In that case maybe we would not need a local KAS (if the widget already connect to a public one) and hopefully we would be able to use Kurento in FIWARE lab. Were you able to make any test in this setup?



3.       Robots:

a.       Will we have more than one robots, also of different types?

b.      Will we have some Robots in Italy and other in  Spain, or all the robots will be in the same site?

Best,
Gianmario


Da: fiware-robotics-bounces at lists.fi-ware.org [mailto:fiware-robotics-bounces at lists.fi-ware.org] Per conto di Iñigo Gonzalez
Inviato: mercoledì 15 aprile 2015 10:09
A: Colombatto Davide
Cc: fiware-robotics at lists.fi-ware.org
Oggetto: Re: [Fiware-robotics] (no subject)

Hi,

streaming video with kurento on fiware lab worked. But we also wanted to try the kurento's video processing capabilities to make smart robots (object recognition, crowd detection, plate reader...) And those examples didn't worked with the kurento's deployment in fiware lab.

We have our own kurento deployed in our machines working with the kurento filters, and we have been working on an app that doesnt' need chrome to stream video tu KMS and use their filters.

Regards,
iñigo

On 14/04/15 18:04, Colombatto Davide wrote:
Hi all,

in the attached file there is depicted the solution that I tried with Kurento some time ago.

The KMS on the FIWARE Lab is an instance started by the Kurento image available on the Lab and updated to the new version of Kurento (as they told to Inigo in an old email).

Inigo, do you confirm that you tried the same solution but maintaining both KMS and KAS on your ET machines?

For me, this solution works but not perfectly: the video/audio has not always a good quality and sometimes the KAS (on my machine) goes down.

As I said I'm not able to use the Trento node because it don't give me a public IP.

Davide
Questo messaggio e i suoi allegati sono indirizzati esclusivamente alle persone indicate. La diffusione, copia o qualsiasi altra azione derivante dalla conoscenza di queste informazioni sono rigorosamente vietate. Qualora abbiate ricevuto questo documento per errore siete cortesemente pregati di darne immediata comunicazione al mittente e di provvedere alla sua distruzione, Grazie.

This e-mail and any attachments is confidential and may contain privileged information intended for the addressee(s) only. Dissemination, copying, printing or use by anybody else is unauthorised. If you are not the intended recipient, please delete this message and any attachments and advise the sender by return e-mail, Thanks.
[rispetta l'ambiente]Rispetta l'ambiente. Non stampare questa mail se non è necessario.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.fiware.org/private/fiware-robotics/attachments/20150417/8e464fd0/attachment.html>