Project Updates for Mu Tag: World's Smallest Loss Prevention Device on BackerKit Page 6

Wed Dec 19 '18 Announcement
over 7 years ago – Wed, Dec 19, 2018 at 06:51:10 PM

Hi Indiegogo Family,

We want to provide a quick update on Mu tag shipping.

400 Mu tag Delivery

This production batch arrived to our distribution partner’s warehouse last Wednesday and all SKUs were tallied into inventory on Sunday. There are two things we are finalizing before pulling the trigger on shipments.

First, we are finalizing the list of final recipients for this batch. While Lang has been straightening out manufacturing yield issues in China, Tom has been busy working with BackerKit and our backer data to manage the logistics of partial shipments. We want to guarantee that everything moves smoothly and there are no mistakes in the accounting of who has received what.

Secondly, Tom has a few more things to to button up on the iOS app. This has been slightly delayed due to Tom managing logistics and product distribution. He finished the data collection system Saturday and now must finish testing and debugging that system. Then we’ll perform some final testing among the informu team and friends to ensure all core features are working as expected.

As long as things continue to go smoothly, product will begin shipping out within the next couple of weeks!

Again and as always, thank you all very much for your support! We couldn’t do this without you!

Mon Dec 3 '18 Announcement
over 7 years ago – Mon, Dec 03, 2018 at 08:18:48 PM

Hi Indiegogo Family,

Everything is on schedule to start shipping out 400 Mu tags within the next two weeks!

We shared a survey with you last update. Don’t forget to fill it out if you want to receive your Mu tags. We can’t ship to you unless you complete it.

What have we accomplished?

Tom received all 400 Mu tag and Mu chargers in the US. Rishabh was able to pitch in and they completed final testing, preparation, and shipped them to the distributor today. We really look forward to our first batch of backers receiving these and would love to hear feedback as we prepare for our next production run. Lang being in China and taking over all work from our previous CM, handling all relations, and personally working with several connector and PCB leads has been a tremendous help. We are able to make improvements quickly across the board as we ramp up production for the next run.

Hardware Production

As you may know from our previous updates, we still had a struggle with low yield due to non working ICs. We decided to go through a 3 tier plan to solve it in parallel fashion. Here is the update on the status of the plan:

Informu’s testing: We planned to do a mini run with new ICs to figure out the issue ourselves. Under Lang's direct supervision, we did two batches of 10 units IC production with our PCBA house.
- We used 10 of the previous CM’s ICs. The IC was previously pre-programmed by their house using their own programmer board before we received it. Lang personally verified that the 10 IC units had Bluetooth functionality before the SMT (using their programmer). He used their programmer to power the IC up. All components are the same as last 900 unit run (where we got 40% yield). This reflow peak temperature was kept at 260’C.
- We used 10 units of the new-in-package IC’s shipped by Raza from South Korea. These were new ICs purchased from Digikey, untouched, and not programmed before PCBA. All components were the same. After SMT, 5 units went through the same temperature profile as last 900 units run (lower than 260’C). 5 units went through with the 10 units IC above together, at 260’C peak temperature.
- After the oven, we went to the x-ray machine to check all 20 units above. None of them had a soldering issue.
- Then, we went to the QA department. One technician used the melter-meter to check if the IC is short, and also powered them up.
- From the previous CM’s 10 IC PCBA, 9 of them had an obvious short, and one IC PCBA kind of lit up briefly (i.e. <10% yield). For Raza’s ICs, all 10 units PCBA were not shorted. It indicated normal current (didn't have any major issues).
- For Raza’s 10 ICs, we programmed them with our original programming method (i.e. after PCBA) through programming pads on PCB and had a good yield.
- Our conclusion: Most likely the 1k IC units previously purchased were bad. The fact that BLE was working on most ICs before PCBA contradicts possibility of ESD damage while shipping as we thought earlier. Also, using informu’s original programming method (after PCBA) showed increased yield when compared to ICs which were programmed before PCBA.
Testing in Boston with Symmetry Electronics: Last week we shipped 3 faulty units to their Boston office where their test engineer will personally test them and give us his feedback on what went wrong after PCBA.
RMA with IC company: Our sales representative from Symmetry is helping us start RMA and thereafter we will ship 5 non-working units to the IC company’s Singapore office to get their feedback as well.

Our next plan is to fully solve this IC issue by doing a mini run using the latest knowledge and checking the yield. We are also working with a connector company to make customized magnetic connectors which will enhance our customer’s experience of attaching Mu tag on Mu charger and help streamline production.

Software

Tom has integrated the machine learning model to improve on false notifications and is finalizing a new data collection system. Once this is complete, we’ll do some internal testing among the team and polish things up as much as possible before the Mu tags start shipping out.

If any major bugs are detected or reported once backers begin to use the app, they will be resolved with high priority. At the point that everything looks good for the iOS user experience we’ll begin finishing up Android.

Delivery

We expect our distributor to receive all product to their warehouse by this Thursday. Then early next week everything will be ready to start shipping out to backers! However, since this is the first time working with this distributor we cannot be sure that an issue won’t occur. We added that extra week buffer just in case.

We are so excited to get the Mu tag into your hands!

Don’t forget to fill out our latest survey if you want to receive your Mu tags. We can’t ship to you unless you complete it.

Again and as always, thank you all very much for your support! We couldn’t do this without you!

Sun Nov 18 '18 Announcement
over 7 years ago – Sun, Nov 18, 2018 at 08:47:31 PM

November Update #1 continued...

Where Are We Now?

We finished all 400 units of Mu tag LED testing, Bluetooth functionality testing, and charging testing. We also finished 385 units of Mu charger LED testing and charging testing at the PCBA house.

Currently, we are in the final assembly process for Mu tags and Mu chargers. It requires us to put the PCBA, battery, and magnets into the plastic parts. Then we use the ultrasonic welding machine to assemble them together. After that, we need to perform LED testing, charging testing, Bluetooth functional testing, and magnet fit testing. It is a very manual process.

Software

Tom has completed all refactoring and testing for the database changes on iOS and fixed several other outstanding bugs. Now he is working on integrating the machine learning model and reduce false notifications. There are probably about two to three weeks left before the iOS app is ready for release. We’ll do as much internal debugging as possible, but will require our backers to put everything through real world use to see if there are other sneaky bugs.

As mentioned in previous updates, the Android app will lag behind iOS and not be ready for the first Mu tag shipment. As soon as the iOS app has been release, then we’ll require several weeks to get Android up to speed. There’s no ETA until the current Android app has been completely assessed.

Delivery

We fully expect to ship 400 working Mu tags and Mu chargers to the United States by the end of next week. Once arrived to the US, we’ll begin packing and labeling for shipment to backers! If everything goes well, we expect to ship out 400 Mu tags within 4 weeks. If we finished the batch in China and ship out this week, then we expect about another week to receive in the US. Then another week to prepare packing and ship to the fulfillment warehouse for final shipment to the backers. An additional week has been tacked on as a buffer.

Don’t forget to fill out our latest survey if you want to receive your Mu tags. We can’t ship to you unless you complete it.

Again and as always, thank you all very much for your support! We couldn’t do this without you!

Sun Nov 18 '18 Announcement
over 7 years ago – Sun, Nov 18, 2018 at 08:43:44 PM

Hi Indiegogo Family,

Wow has it been a crazy stressful, but rewarding last few weeks!

Don’t forget to fill out our latest survey if you want to receive your Mu tags. We can’t ship to you unless you complete it.

When terrible news turns tremendously great

On Tuesday, October 23rd, our Contracted Manufacturer (CM) gave up on our project due to an extremely low yield on the 1,000 Mu tag production run. They had more than a 50% failure rate after Mu tag PCB assembly. Just a few days before, they said everything looked good for producing and receiving the first batch within a few weeks without much risk.

After the shock and panic started to wear off, we decided to make a bold move. Lang Mei, our co-founder, flew to Shenzhen, China immediately to take over the entire manufacturing project. After some negotiation with the CM they were willing to hand everything over, along with their factory partners information in China. We were left with the completed injection mold, 1,000 units of plastic parts at the final ultrasonic welding assembly house, some remaining electrical components, PCBs, 500 broken Mu tags, and other various parts.

The great news is that we have about 400 Mu tags to deliver!

We couldn’t be happier taking over the project completely and working directly with the factories to move things along quickly and with the highest of standards. After meeting with the factories we found out the CM barely communicated with them about our project and left out so many details needed to make a quality product. We have already accomplished more in 3 weeks than the CM did in 7 months. The whole point of the CM was to save the time it would take to build our own factory relationships and provide expertise to save on costly and timely mistakes.

What have we accomplished?

Hardware Production

As we mentioned in the last October Update #2, the 1,000 units of Mu tag PCBs had been produced. It turns out that only 900 units of Mu tags PCBs could be assembled by our PCBA house due to a shortage of the IC inventory. Out of those 900 units of Mu tag, only 402 are working.

This is extremely frustrating to us and the PCBA house. We immediately formed a plan by getting multiple parties involved.

The issue we found was that the ICs stopped working after PCBA, as we had mentioned in our last update. To cater this our CM purchased 1,000 new units and did a trial run and got successful results. Also, Raza hand soldered a few units and they all worked. We thought the issue was solved and were ready for production. It was only after our CM assembled 1,000 Mu tags and got only 40% yield that we realized that we were really missing something. We currently are working with the chip manufacturer “Silicon Labs” as well as their distributor “Symmetry Electronics” who have both agreed to help us debunk this issue. We would especially like to thank Anuja for her constant support and dedication to help us solve this issue of yield.

The most likely issues , after talking with several parties are:

ESD Damage: Currently the ICs are being programmed at another house and shipped to the PCBA house. We think that ESD is likely causing damage either during programming or during shipping. We think so because last year when we did 50 or 100 unit run, we got 99% yield and we only programmed ICs after PCBA. i.e. New ICs were directly fed to the assembly machine without touching them.
Temperature Reflow Profile: Symmetry Electronics suggested we modify the temperature of our oven to 260’C peak and maybe some pins on the ICs (sometimes power pins) are not getting soldered properly and to follow the JEDEC standard.
Programmer Issue: Symmetry Electronics also suggested that maybe the ICs are damaged even prior to PCBA after programming.

Considering these, informu plans to solve the Mu tag yield issue through triple routes:

Ship some non-working units to Silicon labs through RMA for them to figure if the IC is damaged or its soldering issue. We sent them X-ray pictures earlier, but they didn’t find soldering to be the issue.
Symmetry has also agreed to check 3 of our non working units separately with their in house test engineers.
Informu will do a mini run where we follow our original technique and programming method and see if that fixes the yield. Which means that we will not touch new ICs before PCBA and do the programming after PCBA through our in house custom programmer board which we used last year.

All these plans will help us understand if the issue in ICs lie before PCBA or after PCBA. If it’s before, that means ICs are getting damaged at some point before reflow (ESD or programmer damage). If it’s after, then it is happening during reflow (temperature profile or soldering standard issue).

Meanwhile, we are going to deliver as many of the current fully functional Mu tags as possible.

As mentioned in the last update, we introduced some minor updates on the status LEDs from our previous run of the Mu charger PCB . Also, we constricted mounting holes and resized PCB shape to ensure the wiggling issue won’t be there. We quickly did a 10 unit Mu charger PCBA test run. All 10 units worked. Then, we placed a 390 unit order of Mu charger PCBA. So we can get a equal set of Mu tags and Mu chargers.

Vendor Visits

Since we took over the entire manufacturing process, we built new relationships with our vendors by visiting their factories first.

Injection Mold Factory

PCBA

Ultrasonic Welding

Connector Vendor

Continue to November Update #1.2...

(this update went beyond the word limit)

Tue Oct 16 '18 Announcement
over 7 years ago – Tue, Oct 16, 2018 at 02:05:29 PM

Hi Indiegogo Family,

We are back with our another update for October.