Integrating OpenAI with NAO

28 views
Skip to first unread message

Tracy Smith

unread,
Jul 14, 2025, 10:29:18 PMJul 14
to ros-sig-...@googlegroups.com
What is involved in integrating OpenAI LLM with NAO to control motion, language, and APP integration. Please provide a step by step process. 

On Mon, Jul 14, 2025, 6:57 PM <ros-sig-...@googlegroups.com> wrote:
Birhanu Shimelis Girma <bree...@gmail.com>: Jul 14 04:52AM -0700

Hey D.Singh,
 
I also faced similar issue after properly following the naoqi_driver2
readme file( https://github.com/ros-naoqi/naoqi_driver2/blob/main/README.md)
 
But After disabling the audio inside the (
/home/br/ros2_workspace/src/naoqi_driver2/share/boot_config.json)
I don't know why, but It worked after, I also disabled bumper, touch_head,
touch_hand in the boot_config.json file.
 
 
But, the audio thing is still not resolved. Whoever solved it, Kindly let
me know.
 
 
I used the following command to connect to pepper.
 
ros2 launch naoqi_driver naoqi_driver.launch.py nao_ip:=172.29.111.240
network_interface:=wlp0s20f3
 
[image: Screenshot from 2025-07-14 13-43-30.png]
 
 
On Sunday, March 10, 2024 at 6:20:01 PM UTC+2 D.Singh wrote:
 
Vishal Vilas Shendge <vishalvil...@gmail.com>: Jul 14 05:48AM -0700

*Hey everyone,*
 
You can follow the steps I’ve already uploaded to the GitHub repository:
🔗 https://github.com/VishalShendge31/pepper_robot.git
 
If you face any issues, feel free to reach out to me.
 
Happy coding!
Thanks,
*Vishal*
 
On Monday, July 14, 2025 at 1:52:49 PM UTC+2 Birhanu Shimelis Girma wrote:
 
Birhanu Shimelis Girma <bree...@gmail.com>: Jul 14 07:58AM -0700

Hey Vishal,
 
The audio problem still persists while trying to connect to the robot.
Like I mentioned earlier, To have access to all the topics, I have to
disable audio, bumper, touch_hand, and touch_head in the boot_config.json
file.
 
This error persists. Can't find service: ROS-Driver-Audio
 
The project I am working on needs audio for conversation. So is there any
one who has found a fix for it? Thank you.
I am on ROS2 Humble.
vishal shendge <vishal3...@gmail.com>: Jul 14 01:10PM -0700

Hey,
 
You’ll need to set up a complete TCP-based interface where your host device
(running ROS2) sends audio, gesture, or tablet display commands to Pepper.
On Pepper’s side, a Python 2.7 server can receive these commands via
sockets. You can use ALAudioRecorder to initiate audio recording from
Pepper’s microphone and save it as a .wav file. This file can then be
transferred using SCP*.*
 
Feel free to try this approach. If you’re looking for a ready-made
solution, please give me some time—I’m currently quite busy. I plan to
upload the complete solution to my GitHub, hopefully by this weekend.
 
 
You received this digest because you're subscribed to updates for this group. You can change your settings on the group membership page.
To unsubscribe from this group and stop receiving emails from it send an email to ros-sig-aldeba...@googlegroups.com.
Reply all
Reply to author
Forward
0 new messages