uECG - small open source wireless ECG sensor

The wait is over: our first-ever batch of factory-produced uECGs and base stations is here! The package arrived on Monday after spending several days in customs, with assembled devices packed into neat little bubble wraps:

After unwrapping one of the packages, we were in for a shock: never before have we seen one of our assembled devices with perfect soldering, properly-aligned components and no trace of flux. And, as we've soon confirmed after breaking out the pogo pins, they even worked! (by comparison, a manually-assembled uECG requires, on average, 2-3 rounds of resoldering before all the tricky chips work properly)

Here's how they look now:

---------- more ----------

We also never tried pogo pins before, but now that we did, no more soldering wires every time we upload firmware! Only drawback is, you have to hold them down to the board while you're doing it, like this:

We also tested our latest 3D-printed case version, which aids with the holding down by adding holes for the pogo pins:

Of course, not everything went perfect, because the signal was really weird - way off the usual even considering that the electrodes were not attached. There was a brief moment of panic as we contemplated everything that could go wrong, but it turned out to be our mistake - a 510k resistor on the analog side ended up being 510R when we filled out the BOM. Soon as we replaced it, the signal went normal again, so that's problem solved - on this board, at least.

After that, we attached the battery and soldered electrode connectors:

Now, for the other 49 units...

We also tested the base station by soldering the USB connector and plugging it into the laptop, and then seeing if the monitor app worked (it did):

The whole manufacturing process - from the factory quote to new uECGs at our doorstep - took around four weeks, a little more if you count the BOM preparation and changes to the board layout. Now the plan is to solder connectors and batteries to all of the devices, upload firmware, pack them and ship them to Indiegogo backers first before adding the rest to Tindie stock. The packing and shipping part may be harder than it sounds, cause we have yet to laser-cut the cardboard layers to put boards safely for shipping, to order the box sticker prints and then assemble the boxes neatly. That's not counting that we never tried hauling 30 battery-containing boxes with devices to the Ukrainian post office for export.

Eh, better start working on it.

Okay, this is happening - the factory has finished production, and after some more days spent communicating and fixing a led polarity issue (classic), the assembled uECGs and bases are ready. Now we have to wait for them for a few days, then tackle the customs (we never did this).

We produced 50 uECGs and 50 receiver bases. Most of them go to Indiegogo backers, the rest will be in stock for Tindie shop for makers from the waitlist! And yes, we are aiming to start second batch production soon after we ship this one.
Todo stuff for the set assembly and shipping:

- Snap connectors - already here

- Batteries (501240, 250 mAh - more hours than the previous version) - already here

- ST-Link v2 programmers - already here

- USBs for the bases (for manual through-hole soldering here) - shipping from China

- Gel electrodes, screws for the cases, custom pogo pin adapters - shipping  from China 
- 3D printing the cases - need to check the model, then we're ready to print lots of them at a very responsible and accurate guy whom we were happy to find in our hood in Kyiv :)

- Cardboard boxes, cardboard layers, wood stuffing for the shipping packages (no plastic, but have to achieve anti-static properties and damp well) - already tested, need to buy enough for the 50 sets now
- Box stickers design and printing - beta ready, need to design some more and they're ready to print.
All in all, seems like a plan. Still can't imagine us assembling 50 of those.
I guess we'll find out how this feels real soon!

After three days of...

- taking precise measurements of every micro USB, slider switch or tact button in the house;

- filling out proper manufacturer part numbers for every single component down to the last 0.8 pF capacitor;

- browsing lcsc.com to find out that those exact part numbers are either out of stock or 'not recommended for new designs' and then replacing them;

- making several important changes to the schematics and the PCBs on the fly (yay fiducials!);

- finding out the exact reason why those 32MHz crystals are positioned upside down in the library and then flipping them;

- hunting for the common anode on the RGB LED (and then providing screenshots with red arrows pointing out where exactly the common anode is on the PCB);

- briefly contemplating replacing all black chip packages with white ones (don't ask);

- calculating the actual prices of all components on the board;

- and then almost despairing to find out that the .pos file is in the wrong format (minutes before submitting the order);

...we finally sent uECG (and the bases) into production. This has been a few days ago (actually, Wednesday night) and they're still under review so far. Granted, it's only the first fifty devices (each), but that'll be enough to send to our Indiegogo backers and have the rest in stock on Tindie.

There are many articles on the web about the wonders and horrors of PCB production, and almost all of them have one thing in common - that there is no single 'right way' to do it. Just as an example, every manufacturer has a different concept of how their dream BOM should look like. And, most importantly, the .pos file - the one that says where your components should be positioned on the board and how to place them... 

In a sense, preparing production files is fun. It is also hard, exhausting, and time-consuming work (especially the 'find the correct MFN' part). It is also insanely motivating - we're producing our devices, finally! No more manual assembly! And then, at some point, it's over and you check, double-check, and triple-check everything. You upload the files. Hit 'submit'. And then you wait. That's exactly what we're doing right now (waiting), and it's probably going to take a while. 

But we're finally did it! And now that we did, it's time to focus on stuff we put on hold - like writing project logs, picking up other projects (ours and otherwise), and thinking about long-term life goals. Plus, things will be happening soon, so we'll try to keep you updated more often!

After some hard fights with KiCAD libraries, finding pogo pin adapters and last but not least - half of our team moving to another apartment, uECGs and bases for them are almost ready for sending to production. Just need to check the rotation of the components and add some comments to the final BOM.

Time to change that ETA on tindie.

We're finally back and it's been a crazy few days - both on MakerFaire and beyond, as we have never been to Italy and never traveled for such a long time as a team. Saw amazing stuff, met nice people, drank a lot of cappuccinos!

So, in short:

- We're dead tired, but we're starting to get back to life;

- On the way back, they lost one car from a night train we were supposed to take from Rome to Venice (we decided to go the long way), and the station had to arrange a motel for us and the other passengers. The guy on the motel reception didn't even blink when we told him the story - said that's completely normal. We wrote a facebook post about the adventure (and posted some photos of huge Venetian seagulls!);

- By the way, we ate our best pizza in Venice - in Pizza 2000. Good shit.

- Back to the MakerFaire: it was HUGE. Hundreds of makers. Thousands of people. 

The first day, Friday 18th, was Education Day in Italy. Up to 25 000 students boarded the trains from? early in the morning and crowded Fiera di Roma. In a couple hours they ate almost one kilogram of sweets that we brought from Kyiv, and that's only cause the plate wasn't big, and we filled it at some intervals. Well, if that's what it takes for the kids to like technology!

- Saturday was more niche - people were more focused, listened with interest and tested uECG and uGlass. This was also the day when at least two doctors came to our booth, saw uECG, brightened up visibly cause they liked it and said the signal was good - one even gave us suggestions to add a cardioprobe to see the ECG signal from various angles. Very useful, will definitely want to work on that. 

We also learned to explain what our devices did with gestures and simple keywords, as it turned out most Italians don't know English - "Bluetooth ECG, PC, Android, medical signal, electrocardiogram, like in hospital, here - is - from - here (pointing at the device under T-shirt and laptop screen)" and so on. Sometimes just gestures, and it was even faster :)

This was also a very social day - we met Naomi Wu (aka SexyCyborg)! In real life, she's VERY lively, and very friendly. She brought a drone that could charge from a pad with pogo pins and a LED ring - we took some pics of that and a selfie :) She's also stylish and just great, liked her a lot!

In the evening, we went to our first Hackaday meetup! Met a lot of people: Mike Szczys, Larry Bank, Kitty Yeung, Helen Leigh, Drew Fustini (from Beagleboard), Robert Fitzsimmons (he of the large LED), Mastro Gippo (he's the one making PRISM), Laura Gales from UDOO, and other guys! A loud bunch showing each other devices with LEDs in an Italian restaurant. Loved it!

- Sunday was rather quiet. Our table looked the most professional and we answered more clearly after Saturday. Our broken Italian English was perfect. We were getting really tired though, cause we tried to see Rome in between (mostly in the evenings) and slept very little. Also, tending to the booth and answering a lot of people at once is very draining - but it's also something you can't feel anywhere else.

We also took some time to go and see the other booths. There were 10 halls, and we tried to visit at least some of them, take pictures, talk to makers - want to write a separate log about that. There was just so much of everything!

Looks like this is almost it for this log! After we came back and slept for a day, we noticed something's been happening to uGlass page, cause it's been snowballing followers and it's now in top of the projects page. Was that Mike Szczys' comment? :)

Now we'll have to go and write a log there too!

Makerfaire's a big event here: the ads are on the walls, trams, buses and metro turnstiles. Looks really cool, and everybody knows about it, our apartment landlord included.

So with this exposure, we thought, Rome must have lots of makers, and tools and Arduinos sold at every corner. Imagine our surprise when google searches haven't found almost anything about saldatores (italian for soldering iron) when we really needed one (can't take it on a plane) and we only managed to find it after two days of search at some general tools store on the street. Maybe italians are so cool they just stack arduinos when they need to make a device?

Thing is, in our last few days before the travel, we were really busy finishing the commercial projects to send to clients and trying to fit our own projects in any minute left. Managed to assemble three v4 uECGs and USB bases for them, but successfully forgot two uECGs at home and only realized that in Naples. Were really sad about it, but the third uECG v4 still flew with us, so there's that. We also packed:

• a couple older working uECGs
• a new uEMG board and electrodes for it (unsoldered)
• a pair of new DWM1001-based localization modules, and a pair of the older ones
• a working uGlass AR module (attached to glasses)
• a nice round BLE base board with plenty of LEDs, which we've written about in the previous post
• some soldering supplies, wires
• batteries!
• tindie cards, business cards and a poster with the devices and catchy inspirational text about open hardware
• and one kilogram of sweets from Kyiv to give away!

So I guess we still have chances at assembling a nice table. Now we are going to Fiera di Roma by a suburban train – cause the Makerfaire is just outside the city area - to get our Makerfaire badges and prepare the booth for tomorrow, but not before we add some evening Colosseum pictures to this update!

We only got to see it in the evening, and it's so huge. Weird thing is how Italians just go around their business, and it just stands there. Love it here!

A package from PCBWay with a new version of uECGs just dropped! Or rather was brought by DHL courier, which deliver better and better by the shipment - previously we shied away from them, because there were too much red tape and they were arrogant, but maybe they're not so bad after all. 
Anyway, there were three uECG versions before this one:
- the v1.0, made back in December, 2018, which required some wires to work and had ECG snap connectors salvaged from Chinese cheap wired electrodes bought at a taobao acupuncture shop. We found a shop that sold snap connectors after that, which was a great adventure that took some weeks, but we learned some Chinese words along the way.
- the working v2.2 - most featured in videos and tests
- black one, the v3.1 - we had experimented with halvanic skin measurement there, but that didn't work so well. We also planned to add vibe motors to make a gamer feedback device - maybe still do that later.
- and now the new v4 had all bugs fixed (thanks to hackaday commenters for pointing out some of those!): analog and digital ground is properly connected now, and it also has protection in form of a resettable fuse and a BAV99.

We'll assemble them in a couple days and see if they work well.

The second PCB is a multi-protocol base with 16 LEDs and a lot of IOs. It's like LilyPad, but based on nRF52832. We plan to use it a lot.

Like uECG is a simple solution to a usually overcomplicated problem, we have a few other projects that aim to hack modern technology into small, manageable, open-source pieces. Most of the time they exist in different forms - "good idea" seems to be the most popular. But recently a couple of them moved to the next stage.

Back during the campaign we ran a whole battery of tests on uECG to show what it can do. For one of them, we dug out an inMoov arm from an earlier era.
It's not hard to read EMG signal on a ECG-capable device, you just have to use different algos. We have tried a uECG as an EMG-based controller before, but this time there were three of them. Here's a video of the setup in action:

And here's how it actually looked attached to the_3d6's [non-robotic] arm:

We documented this on hackster.io, but the approach had an obvious drawback - it needed three uECGs to work and was a bit bulky. Each uECG processes only one channel, creating additional load that could be streamlined if we had a central controller that could receive signals from several muscles at once. 

That was probably when uEMG was born. Then we designed a board for it, a mix of uECG and a hexagon from the Skulljack project. It also shares some components with it and has 4 channels. 

While it will be able to control an inMoov, we also want to try it for gesture recognition.

At first, the other idea may seem a bit out of our alley. You've probably heard of AR glasses that various companies have been developing for the past decade. Most of them are basically bulky glasses with built-in electronics that's inseparable from the frame. Not to mention, of course, proprietary. We didn't like that. But we also saw an opportunity.

An integrated AR headset is difficult to build and even more difficult to replicate, and also that's more or less what everyone was doing. Instead, we thought it would be nice to have a module that you could attach to any pair of glasses. 

We called the idea - you guessed it - uGlass :) 
Here's a generic OLED module, some optics to project the image and, of course, Arduino, rather permanently attached with hot glue to some glasses:

It just shows its name, battery level and time from startup, but this was enough to get us going. We designed a proper PCB, and on Friday the new boards arrived from PCBWay - here's the assembled modules, fresh from the hot plate:

Now we have to think of a mount or enclosure to hold the optics and keep the module on the glasses. The finished uGlass can be used to show data from several devices, starting from uECG - for example, to display BPM in the upper corner of your vision. The possibilities with AR are endless. If we only have to imagine it being used in medical field, just for displaying basic patient data around the hospital, for example, well... But that's a large topic which we would really like to write about separately!

Originally, we intended this to be just another standard project log.

Then we stopped, examined what we were doing, and wrote this instead.

One week ago, we were super excited about everything going on. Yes, Indiegogo wouldn't release our money, but things were happening all around us - at an alarming rate. We went from nobody to Hackaday Prize finalist (as per our previous update). People have been following our page. We set up our Tindie page and people started pre-ordering uECGs. We had several new devices just about ready for production, with only a few minor touches left (like actually doing the PCB layouts for them). And we were just about ready to solve the Indiegogo problem by setting up a TransferWise account - and then waiting until they do a background check on us to activate that account.

Then everything changed. 

We had to drop everything, literally mid-stride, because of a family emergency of one of the team members - which meant emergency for all of us. Without going into much detail, dealing with that took us an entire week when we could least afford it. Everything - Hackaday, uECG, our other projects, our commercial projects - had to be put on hold until we were done. Once we were done and the situation was resolved, however, everything kind of came back. All at once.

MakerFaire loomed even closer. There were no updates. There were no projects, commercial or otherwise, being done - we had a huge waitlist on literally everything. There was no money - family emergency meant financial emergency, too, as they tend to (and, in this case, some distant relatives from Russia you didn't even know you had). And, on top of that, Indiegogo still wouldn't release our campaign funds.

This was not even Plan B. By yesterday, we were at Plan D: with less than two weeks until we're off to Italy, we'd assemble a handful of uECGs and accessories manually, along with other prototypes we're bringing to MakerFaire. To do that, we'd have to send PCBs into production, wait till they're delivered, and make sure we had all the components to actually assemble them. On the way, we'd also somehow manage to write updates, get back to all our clients we had waiting, and do the project video for uECG to post here (it'll be here soon, by the way). Oh, and we'd also have to eat and pay our rent somehow in the meantime.

Sounds like a plan, right?

Is there a bright side? Well, TransferWise did come through with our account. And we'd continue talking to Indiegogo because at this point, we really need the money. We sent some PCBs into production last week, and they should arrive in a couple days - meaning we'll be able to get to work on them really soon. Everything else was just a matter of time, determination and perseverance in the face of unstoppable odds. In other words, just another normal day at Ultimate Robotics.

Then we realized that... we don't actually have enough components to assemble all the devices. And, belatedly, that today was literally our last chance to order anything from China before the country shut down for National Day week. 

And so we went to work. The kind of work that includes going through all our components stock, taking precise measurements of that one single tact switch we've got left over, and feverishly ordering PCBs at two AM in the night (early morning over in Shenzhen). The kind of work that's usually done like this:

And, at some point, usually includes this:

Was there a happy ending? Kind of. We did what we could. Life keeps throwing emergencies at us, and we don't really have any choice but to keep at it. After all, MakerFaire isn't going to wait for us. And everyone else - we'd rather not keep waiting, too.

Originally, we wanted this to be just another standard project log, the kind where we gloss over most of our daily grind and only leave the relevant bits in. In the end, that makes for a rather dry, if informative, read. 

But it's also something we were frankly tired of doing this time around. 

To all the people who followed and liked our project - thank you. And to the people who pre-ordered uECG - thank you, too. We used to say we couldn't have done this without you, and we still couldn't. When times are the bleakest, it's you who keep us going at it again and again. 

Thank you, everyone! 

We'll be in touch.

Hello - and welcome to our 10th project log! Things have been happening, the short version of which is:

- we've set up our Tindie store, so now you can get uECG pre-ordered there!

- our crowdfunding campaign is over, and we raised enough to find out people are interested in uECG;

- we're going to Rome for MakerFaire, October 18-20, and will bring uECG and a couple of new projects;

- Make:Magazine did a Maker Spotlight on us - here;

- our website is back, and looking better;

- and last, but not least, we (suddenly for us) made it into Hackaday Prize finals! Being busy with our Indiegogo campaign in July, and grappling with Indiegogo bureaucracy through August, we kind of completely forgot about Hackaday - until yesterday, when an email arrived, saying that we're one of the 20 finalists for the Hackaday Prize! This was on top of a wave of activity we had in the last few days, so it's probably time for a short recap of what's happened.

First of all, though, we've decided to set up an online store at Tindie. The store looks a bit raw, but if you want to get your uECG from us, hit this link and it will take you to the product page :)

Second of all, our Indiegogo campaign concluded in July, raising $2031 over the month - our minimum goal was just $600. The campaign has been in inDemand since August, so we've actually raised a bit more money since then, and will use it to produce and ship more devices than originally intended.

Sadly, Indiegogo hasn't exactly been forthcoming: it took us several weeks to get our identity verified, and we've yet to get the funds we raised because of some problem or other. That's why we've been exploring other options to get our devices out there, while also getting ready for production - that's a new one, and we're actually looking forward to it.

Third, we'll be at MakerFaire Rome this October! If you're in the area, come find us at Fiera di Roma from Oct. 18 to 20 :) Rome is a very big event for us, the biggest one in Europe, and we're getting ready for it. We'll be bringing uECG, of course - it's the star of the show. But we also have a few projects in progress that we'll bring as well.

Also, the MakerFaire Rome team got us in touch with Make:Magazine for their Maker Spotlight feature - the interview went out just yesterday, and you can check it out here - we talk about what drives us, what our future goals are and what we believe in. 

Fourth, with all the commotion, we finally got to sorting our website out! Now it runs on Ghost and is still under construction, but with the old website in limbo for so long, that in itself is a fascinating project. 

And last, but not least, the Hackaday Prize - we're honored to have made it into the finals, and we're grateful for the opportunity to participate. Honestly, we never expected that we'd make it. 

To everyone who's liked and followed our project: thank you!