Weekly update: Improved yield results and software crunch time

Weekly update: Improved yield results and software crunch time 1

Happy Friday all!

So, the quick version for those wanting to skip the details are: the yield issues we mentioned last week have improved (yay!) and we are basically working on two remaining hardware issues, the GPS performance and connector. The software and backend are up and running, as we’ve mentioned in the other updates, what’s left there is polishing everything so you’re WOWed from the first use 🙂 Read on to find out more!

Hardware

Our production test, used to test the 100 new PCBs, which seemed to signal yield problems (reported in last week’s update), turned out to be little overzealous and marked perfectly healthy units as suspect. We’ve trimmed the test and the new yield results were much improved, and as a side effect we were able to test the test.

Now that we have a working production flow of PCBs and mechanics, and a full supply of components, only two issues with the hardware remain – the GPS performance and the connector problems to our camera module that we’ve also previously mentioned.

We have performed a massive number of measurements and tunings of the insanely sensitive GPS front-end and the current state is that it does give us position fixes, but not under as many environmental circumstances as we would have hoped. We are not stopping here, we still have a couple of tricks to play, so in parallel we will do more real-world testing and see how it works. It is safe to say that we have learned an incredible amount about radio frequency engineering during this project, something which could be the source for a more detailed blog-post later.

SEE ALSO:  Lifeloggers: Meet the reporter, Amanda Alm

The connector to the camera module has been a board-to-board connector system so far but we’ve redone it to a flex-cable into a lockable connector instead and we are just waiting for the updated camera modules to try it out. Provided that there are no yield issues with this system, the problem can be considered solved but still needs to be tested.

Memoto’s Lund office earlier this week

Software and backend

On the other side, our software developers will be polishing the app and backend during the next two weeks, here are some items from their crunch time to-do list:

Apps

  • Photo sharing – Users can select and share photos to Facebook and to Twitter.
  • Save to camera roll – Users can select photos from a moment and download/save them to the camera roll in full resolution.
  • Delete moment – Users should be able to permanently delete a moment. There should be an “Are you sure?” prompt.
  • Delete photos – Users should be able to select and permanently delete photos from a moment (see delete moment).
  • Notifications when momentification is complete – Users should get a push notification when all new photos have been made into moments.
  • Calibrate play speed of a moment – Make sure the play speed is just right for a user to playback a normal moment.

Backend

  • Star photos support in the API – The API needs to support starred photos.
  • Support for splitting moments in the API – The backend should support moment splitting so this can be implemented in the app.
  • Support for downloading an image in full resolution – The apps should be able to download an image in full resolution.
  • Backend should be verbose – Uploaders and apps need to know what the backend is doing to be able to display information back to the user. For example: the uploader should know if the backend is momentifying photos.
SEE ALSO:  Life with the Narrative Clip: An interview with Ben Stinson

Enjoy your weekend!

/Memoto team