⚜️ ROSCon 2023 Videos and Slides Now Available!

[Katherine_Scott]

ROSCon 2023 Videos and Slides Now Available

Hi Everyone,

I am happy to announce that the videos and slides from ROSCon 2023 are now available online. You can find everything on the ROSCon website and in the OSRF Vimeo Showcase of the event. We’ve also created a short link for the videos that should make the videos easier to share: bit.ly/ROSCon23Vids. To aid discoverability on ROS Discourse I’ve pasted links to all of the talks at the bottom of the post that you can also use.

Along with the videos and slides we’ve put together a quick wrap up post on openrobotics.org. If you want to see some of our professional photos from the event please take a look.

If you missed ROSCon this year and would like a chance to meet up with your colleagues we still have ROSCon Germany next week and ROSCon India coming up in about a month. As for ROSCon 2024, Odense, Denmark is already getting ready for our visit next year.

Finally, I want to thank everyone who made ROSCon 2023 possible! We couldn’t put together such a fantastic event without the help of our speakers, sponsors, volunteers, reviewers, and workshop organizers. You all deserve a round of applause.

ROSCon 2023 Talks

• Opening Remarks
• Building the iRobot® Create® 3 Robot: Challenges and Solutions for ROS 2 Consumer Robotics
• The State of ROS: From the Iron Age to the Jazzy Age
• ROS 2 Types On-the-wire: Type Descriptions and Hashing in Iron and onwards
• Releasing a new ROS 2 distribution
• Panel: Getting Involved with ROS as an Organization
• Licensing of ROS Packages: Changes to package.xml and new tool support
• SDFormat: A robot description format in constant evolution
• Repeatable Reproducible Accessible ROS Development via Dev Containers
• Colcon-in-container: say bye to the “it builds on my machine” syndrome
• ROS with Kubernetes/KubeEdge
• A New ROS 2 Driver for FANUC Robots Utilizing Ethernet/IP
• A ROS 2 Package for Online Cobots Impedance Modulation
• Real-Time Motion Control in ROS: Uniting HAL with Tormach’s ZA6 Robot
• A safe and secure ROS 2 multi-domain architecture for AMD embedded heterogeneous platforms
• Optimizing MoveIt - Costs, Constraints and Betterments
• Leveraging a functional approach for more testable and maintainable ROS code
• Real-time Data-flow extension for ROS 2
• An Integrated Modelling and Testing Architecture for ROS Nodes
• Breaking Bots: Robustness Testing for ROS
• Rclex on Nerves: a bare minimum runtime platform for ROS 2 nodes in Elixir
• On Use of Nav2 MPPI Controller
• Bidirectional navigation with Nav2
• Autonomous Robot Navigation and Localization on 3D Mesh Surfaces in ROS
• ROS 2 powered Android VHAL.
• Lightning Talks Day 1​:zap:
• Up, Up, and Away: Adventures in Aerial Robotics
• NEXUS: A ROS 2 framework for orchestrating industrial robotic lines and cells
• Aerostack2: A framework for developing Multi-Robot Aerial Systems
• Creating scalable customized robotic platforms
• Create® 3 Robot in the Classroom: Teaching ROS 2 to Undergraduates
• ROS 2 On the Browser with WebAssembly for Teaching Robotics
• Lightning Talks Day 2​:zap:
• Octomap is dead: long life Bonxai
• Parameters Should be Boring
• Using Reference System to evaluate features and performance in a standardized and repeatable way
• RosLibRust: Why we built it, why you might want to use it…
• Leveraging Secure Discovery Server in ROS 2
• Simulate your World: A deeper look at extending Gazebo
• Simulate robots like never before with Open 3D Engine
• Simulation of Highly Dynamic Omnidirectional Robots in Isaac Sim
• An Integrated Distributed Simulation Environment weaving by Hakoniwa and mROS 2
• Multi-drone simulation with deep q-learning
• Supporting Robotic Deliberation: The Deliberation Working Group and Tools for Behavior Trees
• FlexBE - The Flexible Behavior Engine: Collaborative Autonomy in ROS 2
• SkiROS2 - A skill-based Robot Control Platform for ROS
• Mastering your ROS system state with Petri nets
• SpatialVision: Bringing Popping-Out RViz to Life with AirPods
• What is Robotics Observability? Scaling ROS from Prototype to Production
• Surviving the Flood (of Rosbags)
• Improving Your Application’s Algorithms and Optimizing Performance Using Trace Data
• ROSBag2NuScenes: Share the Bags, Spread the Joy - Autonomous Vehicle ROS Datasets Deploy
• PostgreSQL / PostGIS to ROS 2 Bridge for Spatial Data
• Efficient data recording and replaying in ROS 2
• Developing on ROS while Scaling
• Closing Remarks

[peci1]

The volume is super low… Could it be reprocessed and made louder? I tried both phone (app) and notebook (browser).

[Katherine_Scott]

Are you referring to particular video? I spot checked some of them and I could hear everything just fine at a very low volume setting.

Are you sure it isn’t your Linux audio drivers?

[peci1]

I checked the intro talk and bonxai. Both have the same level. As I said, I tested on two platforms - Android+BT headset and Ubuntu/Firefox on notebook speakers. In both cases I can barely hear the talk. Listening in public transport is almost impossible with volume up to the ceiling… And yes, other audio works well on these platforms.

Not sure it’s related, but I have similar problems with Sense Think Act on Google podcasts app. Maybe some US-EU discrepancy?

[Katherine_Scott]

Hrrrm . I did a spot check on a few platforms with and without headphones. While the videos do seem a bit quieter than your average YouTube video, I wouldn’t characterize them as unlistenable.

Listening in public transport is almost impossible with volume up to the ceiling

Perhaps this is the problem? I know I have to crank the volume up when I am listening to podcasts on the train and bus. Our inferior Bay Area trains are known to be pretty loud.

In any case, I’ll relay the message to the video production team.

[Ting]

The link of “The Lightning Talks Day 2” is wrong.

[peci1]

@Katherine_Scott

In any case, I’ll relay the message to the video production team.

Thanks. I tried to make some hard data to support my issue. But I’m no audio guy, so it’s probably very amateurish.

Comparison was done with a random video I normally watch and which I consider proper volume (Internet broadcast of Czech public TV).

Kdenlive (top is Roscon, bottom is random video; unfortunately, absolute scale is not available, nor are axis descriptions):

Just eyeing at it, it seems the bars at Roscon are lower.

ffmpeg -i $file -af volumedetect -vn -sn -dn -f null /dev/null returns mean_volume -26.4 dB for random video and -32.7 dB for Roscon. This would mean about a 5 dB difference.

Applying normalization filter in KDEnlive, it outputs this information when analysis is done. I have no idea what it means, but it also seems to be dB of something.

L: -23.606989\tR: 4.411814\tP 0.612335 for random video

L: -28.584723\tR: 12.894199\tP 1.013325 for Roscon

So, if these values are dB, there is again a 5 dB difference in the L item. But I really have no idea

Listening in public transport is almost impossible with volume up to the ceiling

Perhaps this is the problem? I know I have to crank the volume up when I am listening to podcasts on the train and bus…

But I can normally listen to many other content in the same trams and buses Even without maxxing out the volume.

[msadowski]

I just checked the Boxai talk on my end and the volume seems normal to me. It could be some Vimeo trickery, for quite some time it was showing some ROSCon videos as not age appropriate and would not let me watch them. Changing sound levels by region though sounds like a quite odd thing to do though.

[peci1]

It might also be codec-related or sampling frequency-related. But the player offers no configuration options for sound at all.

yt-dlp shows there are many available audio formats, actually:

ID                                                   EXT RESOLUTION FPS │   FILESIZE   TBR PROTO │ VCODEC        VBR ACODEC      ABR ASR MORE INFO
─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
hls-akfire_interconnect_quic_sep-audio-high-Original mp4 audio only     │                  m3u8  │ audio only        unknown             Original
hls-fastly_skyfire_sep-audio-high-Original           mp4 audio only     │                  m3u8  │ audio only        unknown             Original
hls-google_mediacdn_sep-audio-high-Original          mp4 audio only     │                  m3u8  │ audio only        unknown             Original
dash-akfire_interconnect_quic_sep-audio-25c0bcb4     m4a audio only     │ ~ 66.66MiB  196k dash  │ audio only        mp4a.40.2  196k 48k DASH audio, m4a_dash
dash-fastly_skyfire_sep-audio-25c0bcb4               m4a audio only     │ ~ 66.66MiB  196k dash  │ audio only        mp4a.40.2  196k 48k DASH audio, m4a_dash
dash-google_mediacdn_sep-audio-25c0bcb4              m4a audio only     │ ~ 66.66MiB  196k dash  │ audio only        mp4a.40.2  196k 48k DASH audio, m4a_dash
dash-akfire_interconnect_quic_sep-audio-0892f1d2     m4a audio only     │ ~ 23.47MiB   69k dash  │ audio only        opus        69k 48k DASH audio, m4a_dash
dash-fastly_skyfire_sep-audio-0892f1d2               m4a audio only     │ ~ 23.47MiB   69k dash  │ audio only        opus        69k 48k DASH audio, m4a_dash
dash-google_mediacdn_sep-audio-0892f1d2              m4a audio only     │ ~ 23.47MiB   69k dash  │ audio only        opus        69k 48k DASH audio, m4a_dash
dash-akfire_interconnect_quic_sep-audio-050a6651     m4a audio only     │ ~ 34.69MiB  102k dash  │ audio only        opus       102k 48k DASH audio, m4a_dash
dash-fastly_skyfire_sep-audio-050a6651               m4a audio only     │ ~ 34.69MiB  102k dash  │ audio only        opus       102k 48k DASH audio, m4a_dash
dash-google_mediacdn_sep-audio-050a6651              m4a audio only     │ ~ 34.69MiB  102k dash  │ audio only        opus       102k 48k DASH audio, m4a_dash

I tried downloading some of the formats and all of them seem to have the same volume level according to ffmpeg.

And according to ffmpeg, the one of the audio formats is pcm_s16le, 48000 Hz, stereo, s16, 1536 kb/s (default). I downloaded another unrelated Vimeo video with this same audio format and volume is okay (my ears hurt when listening with max volume; Roscon videos are far from that).

---

I must say I’m a bit lost here. I suggest that if you want to debug this further, let’s continue in PMs. If the production team doesn’t have a problem uploading louder videos, that would be the ideal solution for me

[Katherine_Scott]

Sorry, looks like a bug in my script that matches video names to talk names. I went and manually fixed it.

[Pradheep_krishna]

I’m not sure if it’s just me, for some reason I cannot access/download the slides of the talks.

It says:

Not Found The requested URL /talks/Optimizing_MoveIt_-_Costs_Constraints_and_Betterments.pdf was not found on this server.

Edit: It seems to be fixed now.

[tomoyafujita]

@Katherine_Scott

the previous problem happens again? I can see the following error.

https://roscon.ros.org/talks/Building_the_iRobotr_Creater_3_Robot_Challenges_and_Solutions_for_ROS_2_Consumer_Robotics.pdf

[Katherine_Scott]

Try this link

We just archived the page and I think the URLs changed. I’ll need to update the links. It may take me a couple of hours. Sorry about that.