[BUG] turtlebot3 with nvidia profile has port collision with cpu version

[bburda]

Description

When attempting to run the TurtleBot3 demo with NVIDIA GPU acceleration using docker compose --profile nvidia up, Docker fails with a port binding error because both the default CPU service and the NVIDIA service try to bind to port 8080 simultaneously.

Steps to Reproduce

1. Navigate to demos/turtlebot3_integration/
2. Run: ./run-demo.sh --nvidia
   OR
   docker compose --profile nvidia up
3. Observe error

Expected vs Actual Behavior

Expected:

• NVIDIA GPU-accelerated demo starts successfully on port 8080
• Only the nvidia-profile service runs

Actual:

• Error: ports are not available: exposing port TCP 0.0.0.0:8080 -> 127.0.0.1:0: listen tcp 0.0.0.0:8080: bind: Only one usage of each socket address (protocol/network address/port) is normally permitted
• Both services attempt to start, causing port conflict

Root Cause

The default turtlebot3-demo service had no profile assigned, so it started automatically along with the turtlebot3-demo-nvidia service when using --profile nvidia. Both services tried to bind to port 8080.

Environment

• OS: Linux (Docker host)
• Docker Compose: v2.x
• NVIDIA GPU: Required for reproduction

Solution

Assign profile cpu to the default service and update run-demo.sh to use --profile cpu by default. This ensures only one service runs at a time based on the specified profile.