Video SAM2 Semantic Mode Congestion Control for CERN SPS MIRA Robot
Authors/Creators
- 1. Jaume I University
-
2.
European Organization for Nuclear Research
-
3.
Universitat Jaume I
Description
The video demonstrates the use of the semantic mode for network congestion control during an autonomous survey mission conducted by the MIRA Robot at the CERN SPS Accelerator. In this mode, the user remotely supervises the robot from a ground station, relying on a combination of camera feedback and a semantic representation of the environment—including floor, pillars, and pipes—adapted dynamically based on the network conditions. Even under severe network congestion, the semantic mode still enables effective remote human supervision of the autonomous mission by transmitting compact semantic messages of approximately 0.2 KBytes. This approach results in substantial bandwidth savings and a faster, more reliable network response compared to transmitting data from high-bandwidth sources such as cameras, 3D lidars, or laser scans.
Files
2025_07_CERN_SEMANTIC_SAM2_SPS_GUI.mp4
Files
(37.2 MB)
| Name | Size | Download all |
|---|---|---|
|
md5:7bcd8a22220f02d6358d154266733bcb
|
37.2 MB | Preview Download |