we were about to receive our Alpha4 prototype electronics shortly after the last update, where the most important issue was to test and optimise wireless performance. No Aprils fool jokes in this update!
Video is 10m long. Everything I say is written below too.
Why is wireless performance so important?
This is a hugely simplified explanation of the situation to give a high level, transparent understanding.
The right and left Dash "speak" to each other by sending music and data wirelessly through a bluetooth wireless link. The wireless signal can't go through the head, so the bluetooth antennas of The Dash are designed to focus the wireless signals behind the head. We have spent significant time and effort to design the antennas. "Standard" antenna design wouldn't even remotely be sufficient.
To design a great antenna, extensive computer simulations and laboratory measurements are made. The antennas are designed to resonate and perform in perfect harmony with the bluetooth chip and the system it is used with. Important specifications of the bluetooth chip are how loud it can "speak" and how well it can "hear". From alpha3 testing back in January, the "speak" performance was great. The "hear" performance wasn't good enough to enable a solid link (a lot of noise in the system). We thought it was because of noise originating from the dash electronics itself, so we spent a lot of effort reducing the noise emitting from the electronics. All those changes were made to Alpha4 electronics prototype that went into production mid January to be received mid March. When we started testing the wireless performance 2 1/2 weeks ago and the "speak" performance was stellar. The "hear" performance had improved, but was still not good enough. We started deep investigative testing of the antenna with an artificial signal generator "simulating" an ideal bluetooth chip in huge detail, and got results showing that something isn't working as expected from our bluetooth chip supplier.
What is the impact?
In a calm wireless environment (e.g. not a lot of wifi around) the product would work well. As soon as there is more than a few wireless networks, wireless link between the headphones would break up. In any environment where there is a lot of radiated noise, The Dash wouldn't work well.
The sensitivity to noise is caused by the size of our device combined with an issue of the bluetooth chip not performing to specifications provided by the supplier. The manufacturer of our bluetooth chip has provided significant resources to fix the problem, yet not being able to provide a timeline or guidance for a technical fix.
"xxxx" <- place a not very nice word in the x spaces.
This is more than incredibly frustrating.
We always had 2 alternatives in our backhand (plan C & D) to mitigate risk. While we will continue to pursue a solution with the current bluetooth chip supplier and antenna (Plan A) and to improve the antenna even further (Plan B), a solution is out of our hands.
We expect 9-12 weeks for an additional prototype production, software and hardware testing after the bluetooth issue has been resolved. We have weekly jour-fix with our supplier to proactively address the issue. We have no influence on solving the issue, and we have received no indication by when it will be solved by our supplier.
What are the implications of Plan C
As written in the last update, we are committed, capable and have a technological solution (plan C) for finishing The Dash to the specifications presented here on Kickstarter. We will assign significant resources to pursue plan C, to mitigate risk if plan A fails. If plan A evidently fails, we will divert all resources to plan C.
Plan C a significant technology pivot that will cause a delay.
A significant benefit would be additional battery life to a theoretical 4,5-5 hours while playing back from The Dash embedded music storage. Major disadvantage is only on BRAGI's side. The solution is considerably more expensive to build than plan A.
Timeline of Plan C (manufacturing)
Hardware prototype (Alpha5) by mid May
Hardware protoype (Alpha6) first week of July
Developer production run (Beta) end of July
Pilot production (1000 units/October) last 2 weeks of August
Mass production (2000 units) second week of September
4000 units third week of September
10000 units fourth week of September
Building The Dash
By the time you will receive The Dash, it will consist of 521 individual pieces put together in a tiny package. This includes plastics, LED's, connectors, stickers, manuals, screws, gaskets, stamped metal, light-guide material, sensors, capacitors, resistors, silicon lenses, 7 pcb's, clips, inductors, magnets, cables, microprocessors, memory chips, fabrics, clips, coils, amplifiers, membranes, filters, level shifters, protection circuits, batteries, packaging, foils etc. from 31 different suppliers.
Every single component has been cost negotiated with suppliers, and the suppliers are selected and awarded this and next week. The product will be produced in China by a contract manufacturer that we have worked with for over a decade.
Everything is set up to push the button. Trust me, I've been dreaming about that day for the last 2 1/2 years!
Injection mould tools
Most of the tools already exist and have been reviewed. We mainly need a few modifications to the existing tools and a few new tools. We are very happy about progress.
Testing of The Dash
Every time a developer releases code to our system, that code has to be tested against our user experience criteria and towards automated tests qualifying that code.
David, Torsten and Toby have been significantly expanding on the capabilities of our testing setup. By the time that we receive Alpha5 prototypes, we will be able to have 50 pairs of The Dash in our test rig, that can be assigned to parallel test routines.
We should be receiving our environmental test chamber (cold/heat and humidity) next week, to perform accelerated lifetime tests on the electronics.
The test setup is done to ensure as few bugs and failures as possible when we launch The Dash.
We showed a clickable version of our app a few updates ago. We used the feedback from you to further improve the usability. The app development team (Martin, Luigi, Michi, Arne, Rene and Christian) are working on iOS, Android and Windows Phone versions.
Open house day @ BRAGI
40 people go to work every day in our office in Munich. We all think that we are on to something special, and we would like to share that feeling with the ones that made it possible for us.
I'd like to invite our backers to our office in Munich, Germany on the 13th of May at 4pm (15 people maximum due to fire regulations) for a tour and a beverage at the local bar. We will not prepare any special demo's (delivery of final product above everything), just give you an overview of our work, see The Dash in real life and be able to chat with our developers.
We will make a small video to be sent as an update after the open house day to share a bit of the experience.
To attend open house day, write a mail to firstname.lastname@example.org
Do please only write if you really can attend. We will randomly pick the 15 attendees by 13th of April.
2 incredibly lucky backers
We will randomly pick 2 people from the BRAGI backers who comment "I'd like to meet BRAGI" in the comments of THIS update.
BRAGI will buy a (return) economy flight ticket and 2 overnight stays in a local hotel for the event on the 13th of May (all other costs are not covered by BRAGI). "I'd like to meet BRAGI" has to be written into the comments and be visible to BRAGI by 4.00pm CET 13th of April 2015 to be acknowledged.
We will announce the winners in the comment section and contact them to arrange flights.
BRAGI TEAM and Nikolaj
BTW, the few who requested a refund have been swiftly processed throughout the last months.