The Mark II is Mycroft AI's second smart speaker product. It is a helpful smart speaker running Mycroft AI's own Voice Assistant software. The Mark II can answer general questions, play music, tell you the weather, set timers and alarms, and much more. Mycroft AI is focused on creating privacy-respecting voice assistant products both for consumers and through B2B2C partnerships. Mycroft is also an open source project, much like Android, Linux, and much of the software used to run the modern web. Mycroft AI's open source approach allows it to greatly expand the number of developers involved in the project and create a dedicated and passionate group of users.
After completion of the Mark I project we realized that there were several flaws that we wanted to address in a version two product. The biggest issue was the Audio Front End for the Mark I was insufficient to do great Barge-In, which is the ability to interrupt the Voice Assistant while it is playing music or speaking. Secondly, we wanted the audio quality to be much better and more competitive with other voice assistants on the market. On top of this we decided to move from a limited dot matrix display to a full high quality LCD for additional information. We were able to successfully deliver on all of these goals.
This is a blocking model of the initial components for the Mark II. This was a collaborative process with our hardware development partner, Kevin Jordan of Netkruzer Technologies. There were many constraints to the design. The Mycroft custom PCBA needed to connect to the Raspberry Pi 40-pin GPIO connector. We decided to connect the two boards together at right angles to allow for better airflow. Working around the set placement of components on the Raspberry Pi 4 Single Board Computer was a challenge.
