How to Build An AI-powered Transcription Bot for FM Transmissions

Presentors: Brian Callahan AD2BA, Zhemin Zhang KD2TAI
Moderator: Vince d’Eon VE6LK

Collateral or Slides:
FMBot - QSO Today March 2023 Brian Calllahan AD2BA and Hisen Zhang KD2TAI.pdf

Level: General Audience

Category: Controllers - Raspberry Pi, Arduino, Future of Amateur Radio, Homebrew and Construction, New License - Now What?, Software and Satellites, General Interest, New License- Now What?, SDR Software Defined Radio, Software and Services, Station Operation

Time:

PDT: March 25, 2023, 4:00 pm
UTC: March 25, 2023, 11:00 pm
Duration: 1 hour

Based on the January/February 2023 QEX article of the same name and subsequent ARRL-TAPR conference paper, this presentation discusses and demonstrates a bot, which can be built using as little as an inexpensive SDR antenna, a Raspberry Pi or laptop, and open source Artificial Intelligence/Machine Learning software, that can listen to VHF/UHF frequencies for amateur radio transmissions and transcribe those transmissions into text. We will discuss and demonstrate potential uses of the transcribed text and the bot generally. The bot can augment anyone’s experience and enjoyment of being on the air, from new operators who may only have their Technician license to seasoned operators.

SDR Rover Kit: The Next Generation Setup for Amateur Radio Contesting

Presentors: Zhemin Zhang KD2TAI

Collateral or Slides:
SDR Rover Kit - QSO Today March 2023 Zhemin Zhang KD2TAI.pdf

Level: Intermediate
Category: Contesting and DX, Future of Amateur Radio, Homebrew and Construction, Field Operation, SDR Software Defined Radio

Time:

PDT: March 26, 2023, 10:00 am
UTC: March 26, 2023, 5:00 pm
Duration: 1 hour

The SDR Rover Kit project uses off-the-shelf software-defined radio for an 8-band low-power contesting solution under $400. The idea is to keep the system “simple and stupid” by using wideband components. The prototype has been proven effective in lab and field tests. However, it takes additional optimizations to achieve ideal performance.