AI could detect network instability by tracking RCS-to-SMS fallbacks

Telecom systems engineer Srinivasa Maringanti proposes using RCS-to-SMS and VoLTE fallbacks as signals of hidden mobile network instability. AI models could find patterns across billions of messages and tie them to specific locations, which aggregate performance metrics tend to hide.

• RCS-to-SMS fallback is invisible to users but hints at network problems
• Signals sit at the individual message and call level, impossible to monitor manually
• City-level average metrics hide issues in specific buildings and neighborhoods
• Models are trained on open data mapping RCS events to SMS and MMS