Which Aragon is the optimal choice for your personal solution?
What is the minimal Aragon system in terms of devices and cost?
What do Aragon1 and Aragon2 have in common?
- Aragon1 and Aragon2 are offline voice control systems for Home Automation.
- Both systems respect user privacy and do not require an Internet connection or cloud services.
- The hardware of the Aragon1 and Aragon2 Satellites WL and PoE are identical.
- Aragon1 and Aragon2 use the same ASR and NLU: Kaldi.
- Aragon1 and Aragon2 come with Node-RED, pre-installed and configured.
What are the differences between Aragon1 and Aragon2?
|Operating system: Debian 9 (stretch)
|Operating system: Debian 11 (bullseye)
|Basic TTS but offers online Amazon Polly
|Enhanced built-in TTS
|Training and configuration of ASR/NLU less flexible
|Training and configuration of ASR/NLU customizable and more flexible
|Web UI in English only
|Web UI in German, French, and English
|Core features are developed and licensed by Snips, now Sonos. The software is no longer maintained.
|Software does not depend on third-party licensed software for speech recognition.
|Software is actively developed and supports features like Homebridge, Netdata, Tailnet VPN P2P, and optional online transcription and AI services (e.g. OpenAI)
|Powerful and designed to remain relevant in the future: 1.8GHz 4+2 CPU, 4Gb RAM and 32Gb eMMC
|One wake word ‘Hey Snips’
|Wide selection of wake words, including ‘Hey Snips’
|Offline Kit available for isolated networks
| Easy-to-use interface for configuration of the ‘all-KNX’ gateway
(KNX via NetIP)
What do ASR and NLU mean?
ASR stands for Automatic Speech Recognition. It is a technology that converts spoken language into written text. ASR systems typically use a combination of an acoustic model and a language model, and they often employ machine-learning techniques.
NLU stands for Natural Language Understanding. It is a sub-field of artificial intelligence that focuses on machine reading comprehension. NLU models interpret the text produced by ASR, providing structured data that a system can act on. This process allows the system to understand and respond to voice commands.