Blog

Manufacturing Updates

Dear Reader,

The last week, like the week before it, was a combination of manufacturing and app development.  It’s pretty clear now that getting the mobile apps finished and in the Apple App Store/Google Play Store will be tasks driving the schedule.  Unfortunately the Mooshimeter is not useful without a mobile app to talk to it.

Why not release the API and crowdsource the apps?

In short, because the app development process is teaching us a lot about the different ways iOS and Android implement Bluetooth Low Energy (BLE), and that drives changes to the API.  BLE is designed for very low bandwidths, usually a few bytes per second.  To quickly stream across graphs like you see in our videos takes some “protocol abuse”. This is where the real differences in the iOS and Android implementations of the protocol are exposed.

After establishing core functionality I hope that users will expand upon/write their own apps.

Racks on Racks on Racks

In the last update I showed you the calibration and testing stands.  We’ve run a few hundred meters through the process and done final assembly on almost 200 of them.

First fully assembled, programmed, and calibrated rack of meters!

More assembled meters. Getting a little cramped.

20141027_220510-MOTION

Installing fuses before calibration

Racks of electronics awaiting calibration

 

On new meters, we’re seeing an average uncalibrated accuracy of about 1% with 0-50mV offset at the ADC.  This is good.  By design the uncalibrated tolerance should be about 2%.  Presently any meter that shows an uncalibrated accuracy worse than 2% off is flagged and set aside.  We might accept them later, but I want to get a better look at them and make sure nothing insidious is going on with them.  FYI only about 5% of the meters so far fall outside of 2% uncalibrated threshold.  Immediately after calibration the meters are well within 0.1%, I’ll check in on the calibrated meters again regularly to see how that drifts with time and temperature.

“I’m a Mooshimeter! Connect to me!”

The Mooshimeter is always on.  When a smartphone is not actively connected, it sleeps to save power and wakes up every few seconds to send out a radio packet advertising its existence.  It’s just a little “Hey, I’m here!  You can connect to me!” message to anybody who might be listening.

When you have 200 Mooshimeters all next to each other, that little “Hey, I’m here!” message once every few seconds becomes an overwhelming torrent to any radio listening on the right channel.  There are other offices in my building that work on Bluetooth Low Energy devices, and now they must filter out a constant torrent of Mooshimeter broadcasts.  This might be a bigger issue down the road, but for right now the kind folks at Sheepdog Sciences are just being nice about it.  Thanks guys.

What’s next?

My main focus now is getting all the core functionality in the apps working.  In the next update I’ll try to put together some screenshots.  Final assembly and calibration are happening simultaneously, but I am less worried about them because the time required for those tasks is easily characterizable.  Estimating the time required for software development and debugging tasks is much harder.  If you’re interested in the topic of why estimating software development time is hard, I encourage you to read this excellent essay..

As always, feel free to email or comment below.  Thanks for reading,

~James

21 Responses to “Manufacturing Updates”

  1. AMS November 4, 2014 at 6:46 pm #

    Will there be a Linux client / library / pile of python code at release?

    • James November 6, 2014 at 2:40 pm #

      Not at release, but the BLE profile will be open so you should be able to write your own. I have some hacked up python scripts that talk to early versions of the meter on linux through a TI USB dongle, but it would take some work to make them releasable. Priority right now is iOS and Android. And getting the meters themselves done as well, obviously :-)

  2. Marek Ištvánek November 5, 2014 at 4:25 am #

    Good news. Good luck with the software development.
    I know what You are talking about as a software developer as well :-)
    I am looking forward to incorporate the Mooshimeter to my SenseLab Windows application I am developing now…

  3. Beat November 5, 2014 at 11:56 am #

    Thanks for the update!

    Don’t forget to remove the battery or put a plastic strap to insulate it for shipping.

    Customs and Airlines do not like equiments which are on and sending, and might return them. That happened to the Pebble Bee KS team!

    Btw, you should really privately contact them for BTLE experiences exchanges. That might save you huge development time. Check out their updates on KS, they went through BIG pains (e.g. IOS 8, including 8.1.0, has introduced new bugs that did not exist on IOS 7, and 8.1.1 might not yet fix them. And Google is making big progress catchups compared to IOS 7, as Android 4.4- BTLE support was random. And Android 5 seems to finally solve the issues, according to their updates).

    • James November 6, 2014 at 2:19 pm #

      This comment is gold. Thank you! I had previously checked with USPS and they said they had no problem with “on” equipment as long as there’s no lithium-ions involved. I was planning on shipping them “on”, but now must reconsider. I don’t want to require the user to open the case upon receiving the meter…

      I’ll reach out to the Pebble Bee folks. Maybe they will have some insights for me.

      Thanks again :)

  4. Jim Twitchell November 5, 2014 at 2:14 pm #

    whats the expected battery life? Given I will be an occasional user of the meter, I wonder if it would be smart for me to remove the battery between uses…

    • James November 6, 2014 at 2:09 pm #

      Sorry – the way the blog is set up a new user’s comment has to go through moderation. Now that you’re all registered I think your comments will appear without moderation.

      I haven’t recalculated battery life in a while. Last time I did I think I got about 1.5 years in standby.

  5. Boyd November 6, 2014 at 11:41 am #

    Thanks for the updated information. So will the meters ordered ship before the end of November? What further bugs in the firmware and software do you need to work out. I understand it is difficult to say but you should be very close to releasing the product?

    • James November 6, 2014 at 2:21 pm #

      I still think it can happen in the next few weeks… the path forward is visible, it’s just not clear how fast it will go. Yes, I think they can go out before the end of November. And if that changes you’ll find out here first.

  6. Achim November 6, 2014 at 5:05 pm #

    Beats comment is absolute correct. Airlines and freight companies will carry perhaps one part with batteries active but if you like to transport many of them they won’t.

    • James November 7, 2014 at 7:09 pm #

      Thanks Achim,

      I received confirmation from yet another source that the post office doesn’t like lots of radio transmitters. *sigh* Expect a blog post soon.

      • Thomas November 8, 2014 at 2:05 pm #

        Before you open them up and start disconnecting the batteries… this requires a software hack.
        The meter could be instructed (over the air) to enter a “shipment state” which doesn’t talk.
        It could recover only when the voltage terminals are connected to a few volts for a several seconds. This wouldn’t be an unreasonable burden for the new owner.

        • James November 9, 2014 at 3:37 pm #

          You beat me to it :) This is almost exactly what I ended up doing, except instead of requiring the meter to be hooked up to a voltage source, the meter will wake if you short out the resistance measurement for 10 seconds. When that’s done, it writes a flag to flash and never re-enters shipment mode. I’ll write about this in the next update.

      • Beat November 8, 2014 at 7:57 pm #

        Maybe a very thin flexible plastic stripe that can be pulled out would do it for now ?

        Obviously in version 2, an on-off switch would be good to have. Some people (or their wifes) don’t like radiations all the time (even if BTLE is really low power), for various more or less valid reasons: health, security, or developing a BTLE device where you don’t want noise traffic.

        Or should the Mooshimeter be silent by default (listening mode), and only reply when asked to communicate by your app ? I’m not familiar with BTLE if that’s possible. But that could be a solution…

        • James November 9, 2014 at 3:42 pm #

          External controls are troublesome because of the safety tests you need to pass to be a CATIII meter. There’s a reset button on the Mooshimeter, but now it’s buried behind the plastic enclosure because it just barely didn’t pass the safety test. In the back of my mind I’m thinking a capacitive touch control could do the job, but that would obviously have to be another version. One thing at a time :)!

          Regarding meter defaulting to silent mode: Unfortunately that’s not really a power-viable option here. “Scanning” actually takes much more energy than “advertising”, because you have to hold the radio on for an extended period of time. When advertising the MCU just has to wake up, blast out some packets, listen for a response and go back to sleep. The whole process takes roughly 5 milliseconds.

          • ivan November 11, 2014 at 1:21 pm #

            It probably isn’t possible with the current version, but with the transparent case, light would make a nice wakeup trigger. Considering it is not a good idea to work with electricity in the dark and the included carry bag is a nice place to keep the meter, which is also dark, it would be very nice and comfortable to have a sleep mode, where it only wakes up (temporarily) when there is (some) light. Your resistance shorting standby could be a reliable silence mode for other use cases.
            Regarding silence, I might not want everyone around to know I am carrying/storing a mooshimeter … at least until I open the bag :)
            As for battery life, with the current standby expectations, everyone will need to take the meter apart and replace the battery at least every 1.5 years, e.g. when they need it, after a few months of storage, and they might not have spare batteries. Lower self-discharge e.g. mini-photovoltaics powered wakeup/standby would be a nice feature. Nice unimportant feature :) For those who won’t be using the meter constantly.

      • AMS November 9, 2014 at 2:08 pm #

        Maybe have it keep the radio off until voltage is seen on the voltage terminals. Check them at 1Hz or something? That way no need for opening the case or a pull tab and you could leave them powered, but quiet until delivery.

  7. Nathan November 9, 2014 at 1:53 pm #

    I would be very interested in your ios experience and ble code snippets as I have started to teach myself ios/swift with a view to interact with micros.

    • James November 9, 2014 at 3:53 pm #

      Then stay tuned :)

  8. Rob November 12, 2014 at 1:54 pm #

    I have experienced the maddening process of “things not going as anticipated ” myself. It sounds like you have the right attitude of slugging your way through it. Keep it up, and don’t get discouraged. The trials you face and overcome are simply tuition towards success.

  9. Boyd Barrie November 19, 2014 at 12:06 pm #

    Is there any update when the meters will ship. I noticed a charge on the credit card.

Leave a Reply