I wanted to start this update by apologizing for the company not communicating through Kickstarter and Indiegogo as well as we should have.
We send out a newsletter once a week about releases and progress across Mycroft. If you don’t already receive this, please sign up at https://mycroft.ai/newsletters/. Progress across Mycroft’s range of technologies is critical to releasing a top-quality Mark II device. Without a reliable, high quality software stack to provide the user experience, the speaker itself is a paperweight. However we have also heard that we need to do better at communicating on how this relates to the Mark II, and on the hardware progress itself.
Updates to Kickstarter and Indiegogo will now be limited specifically to Mark II updates. There will continue to be at least one update each month, but if you want more regular updates, sign up to the newsletter, or join the Mycroft Community Forums and Chat.
To date we have built 15 prototype devices based on the Raspberry Pi 3, a Respeaker Mic Array v2.0, a 20-watt Adafruit Amplifier, a stereo loudspeaker, and a 4.3” Waveshare display. As we have mentioned before these specifications will not be the final production design, but they are acting as the absolute minimum representation of the resources we will be using in the production device.
You may be wondering why we are we still using Raspberry Pi 3s for our prototypes. The answer is simple, we are able to make significant progress on the software on the stable Pi 3 without having to troubleshoot the issues that arise when using the newest single board computers. We have experimented with Raspberry Pi4s and competitor SBCs but the challenges around support, power and heat have always led us back to the Pi3. This focus on stability and performance means that when we switch to the production platform our software will run optimally. If it runs great on a Pi3 it will run even better on our production hardware.
You can see the latest prototype running below. We’ll continue to post these basic run through videos with each update to show progress of the entire experience. Technical details are important but at the end of the day, humans like yourself will be using the device.
We learned some hard lessons from the engagement with our previous hardware partner that I’ve talked about previously. Making complex electronics is hard. Really, really hard. Global supply chain management, uncertain development timelines, lying vendors, low quantity fabrication, quality control, firmware development, software hardware integration, intellectual property management, personnel recruitment and retention, fundraising, public relations – all of these have to be handled and, in the case of a startup like Mycroft, it needs to be handled without the billions of dollars in cash and human resources the giants of Silicon Valley have at hand.
Whilst we got burnt in the past, I also know we cannot do this alone. So our team has been meeting with several well established companies to do the production engineering for the Mark II hardware. The leading candidate to build the Mark II is extremely experienced in building hardware voice assistants. They have delivered products for top tech companies and startups alike, and moving forward we know they are capable of helping us deliver a high quality product. We will be able to formally announce this partnership shortly, once we have a signed agreement in place.
I won’t speculate on a delivery date, however that doesn’t mean we don’t have timelines we are aiming for. Our design team has developed a detailed timeline for production based on the meetings we’ve had with the hardware production company.
This is not an ideal timeline, but it is a realistic one. We’ve heard your frustrations at the delays, and I can promise you that our team has been frustrated too. If we could ship tomorrow we would, however we need to get this right, not just shipped. This timeline is what it will take to produce a top-quality voice assistant that we can all be proud of.
At any time, please feel free to contact us at email@example.com.
CEO of Mycroft A.I., serial entrepreneur and one of the few entrepreneurs in the United States to build a gigabit fiber network from scratch. Joshua brings more than 15 years of entrepreneurial experience to the Mycroft team.