Bluegiga, OHS, Maker Faire, and Seeed Studio
As most of you know, I got to participate in the Open Hardware Summit and then World Maker Faire at the New York Hall of Science at the end of September. This was an excellent opportunity, and for those of you who remember my update about the Keyglove exhibit at Maker Faire last year, let me say that this one went significantly better than last time--no sickness, no loss of voice, much better preparation, and of course more development done on the project. OHS was great as well, but I'll explain more about that in a moment. First, a bit of news about why it has taken me a full month to update you all about the trip to NYC.
In the middle of August, I had a fantastic opportunity fall into my lap: Bluegiga Technologies offered me a position as the senior Field Application Engineer for their US customers. Bluegiga is the company that makes the WT12 module I use in the Keyglove for wireless communication, and it was entirely due to my search for a good Bluetooth module for the Keyglove that I found them in the first place back in 2011. I was just a customer then, and their products, documentation, and incredible customer service got me hooked.
My subsequent efforts to create a few breakout boards for my own prototyping (and later resale) and the Arduino code library I wrote to interface smoothly with their iWRAP firmware put me on their radar, and so they came to me a couple of months ago with a very good offer, which I accepted. It would have been crazy not to under the circumstances, though I and my previous employers were sad about my departure from their company--I've been a web application developer there for almost 10 years, since before I started college, and I enjoyed both the people and the work. But this position at Bluegiga was too good not to take.
Fast forward to the beginning of October, literally one day after Maker Faire ended. I had a week-long trip to Atlanta for introductory training at Bluegiga's US headquarters, followed immediately by two weeks of more focused training at their main office in Finland. I got back from there on Sunday morning (the 21st of October), and I actually wrote a good part of this update while sitting on a plane on my way back from another week-long event in Chicago! The schedule has been crazy, but it's settling down a bit after this last trip, and I should be able to get back to a normal routine soon. There's still a whole lot of steep learning curve for me to overcome in a very short time.
So what does this mean for the Keyglove project? Well, it's still mine to work on; changing day jobs doesn't affect ownership. There is no competition with Bluegiga's products--quite the opposite, in fact--so they are more than happy to have me continue working on it. Most importantly, my head has now been crammed full of knowledge about their products, including some of their upcoming modules. I will be able to make the WT12 integrate as well as possible with the Keyglove, finish the iWRAP code library (which is still missing a few features), and plan much more thoroughly for incorporating Bluetooth Low Energy in the future. This will allow for better integration with new iOS devices and any other new hardware that supports BLE.
In short, my new position at Bluegiga will help tremendously in the areas of Keyglove development that use wireless technology. I am really excited about this.
Open Hardware Summit
The Open Hardware Summit on September 27 was the first main event that we attended in NYC. (By "we" here, I mean myself, my wife, and her mom who came along this time and helped out a ton--thanks, Sonia!) The event took place at Eyebeam in the southwest part of Manhattan, and held at least a few hundred people. There were a few dozen speakers and even more exhibitors, including Super Awesome Sylvia, for whom I had the pleasure of demoing the Keyglove that evening. I met some great new people and a few current friends, and learned about amazing tech and even some things that will help with Keyglove development. One of the guys from Mach 30 even did a quick interview of me at the booth, which you can watch here if you're interested. (Pardon the noise; it was an incredibly loud environment.)
I most enjoyed the OHS speakers for the day, including a great keynote by Chris Anderson of Wired Magazine (his presentation starts around 11 minutes in on that keynote link), a humanitarian Geiger counter project by the guys at Freaklabs, and some interesting discussion about intellectual property and licensing when it comes to open-source hardware. One of my favorite lines: "Give away the bits, sell the atoms." Bingo! Chris Anderson's tale of a Chinese product cloner converted into a value-generating contributor was another great lesson. Structuring a company and development process to encourage openness is definitely the way to go. Or, at least, the way I'm going.
One of the main Keyglove lessons I learned at OHS is that the sensor wire connections still need work. They are easy to solder now that I have metal contacts, and electrically they are quite good. But mechanically, they are far too easy to break after normal usage by cold-working the solder joint. This means that I need some better strain relief. This could come in one or more forms: one guy recommended that I simply crimp the sensor wire in place between the two layers of the fabric snap, and avoid soldering altogether, since this is what made the stranded sensor wire so susceptible to cold-working. I really like this idea, but I need something that will also work for repairs, and there's no way to re-use a fabric snap to crimp a new wire end in. You'd have to replace the whole snap...which could work, but is quite difficult to do once the glove is already built.
Another option is to use some kind of plastique glue, like a metalized epoxy or something with a silicone component. Depending on the conductivity of the glue itself, this could be used either instead of or in addition to an actual solder joint. One way or another, I have to improve the sensor connection technique, since at the moment it is the primary point of failure, and there's no way I would feel good about distributing something that is prone to break so relatively easily.
I got an OHS album up on the Keyglove Facebook page if you would like to check it out.
Maker Faire NYC
This year, Maker Faire went fantastically better for us compared to last time. As I mentioned above, none of us got sick, I didn't lose my voice, and we got to stay and exhibit both Saturday and Sunday. I had a demo video looping on two iPads the entire weekend to help save repeated explanations, and that turned out very well. (That video is almost ready for publishing online, but needs a couple minor adjustments.)
We didn't run out of any materials, despite the fact that easily over a thousand people came by the booth over the course of the weekend. Giving away buttons for people who typed their names with the Keyglove was a fun addition this time around, and I heard from some people way over in the Maker Shed later that weekend that they saw a lot of people wearing them. Score!
Having multiple gloves was definitely helpful for accommodating those whose hands were not the same size as mine, and having multiple computers made it easy to run a personal demo from behind the booth at the same time as visitors got to try out the prototypes. At least three people said that it was the coolest thing they had seen at Maker Faire--which I know is very subjective and depends on what order they saw stuff, but it was still amazing to hear.
There were many people who asked for an explanation about what the Keyglove is really useful for, but unlike last time, I felt like nobody at all walked away from the booth unconvinced of the possible utility in at least one relevant area. Some liked it for the novelty, but pretty much everyone was excited about medical possibilities, or gaming possibilities, or 3D design input control applications. Somebody even mentioned that it could be useful for live, detailed feedback from a patient inside an MRI machine, since it could be far enough outside the machine and requires small enough movements not to disrupt the process. Who knew? This influx of new ideas is one of my favorite parts about exhibiting.
One of the other highlights of the weekend for me was getting to meet with Zack Freedman, who had a couple of weeks earlier finished a working prototype of what he calls a Pi-Borg, which is a combination of a Raspberry Pi and a modified MyVu Crystal EV head-mounted display built into a wearable computer. He stopped by the Keyglove booth, and we hooked up one of the Keyglove prototypes for what was the very first realization of my personal target use case for the project: a fully wearable computer. It was clearly a rough draft since there were some immediate calibration issues with mouse control, but as a successful proof-of-concept exercise, I got excited like you wouldn't believe. Well, actually, most of you probably would believe it.
On Sunday, John Baichtal of Make Magazine even gave us an Editor's Choice award for the exhibit! Thanks, John!
I got lot of good feedback over the weekend and met some great people, and I can hardly wait to get back to the project in earnest now that I am finished (for the moment) with all of this new job-related travel.
Check out the Keyglove Maker Faire photo album on Facebook if you are interested.
Speaking of getting back to the project, I have some good news about some components I have been able to get from the excellent people over at Seeed Studio and the suppliers they work with directly. I expect I'll be able to use their services even more as some of the last pieces fall into place for the kit version of the Keyglove, and then even more for sourcing, fabricating, and assembling the customized all-in-one controller boards. Despite being 18 months out from the end of the Kickstarter campaign, I still have a sizable chunk of the funds you all provided, and with an outfit like Seeed, it will actually go quite a long way. They are a great team to have available, especially for individual hobbyists like me.
Anyway, in the weeks prior to Maker Faire, I worked with Seeed to get some pre-fabricated rainbow header cables and molded (not cut) female headers to go with the Keyglove Kit PCB. Preparing the sensor cables was the most time-consuming step, after attaching the touch sensors to the glove itself (still working on that one).
As it turns out, getting those cables pre-made from Seeed not only saves a boatload of time, but the materials cost through them is actually far lower than what it costs me for the individual components! Amazing. The same thing goes for the female header pins, which are surprisingly expensive otherwise. In short, what had required about $10 and two hours (for each glove) now requires about $3 and no time at all. I will definitely keep employing their services as long as I have needs they can fulfill.
So, what's next, exactly? First, especially while everything is fresh in my mind, I'm going to go back and wrap up the wireless functionality with the WT12 Bluetooth module. This is a longstanding incomplete feature that will be good to have all finished. This mainly involves updates to the iWRAP code library I built and integrated into the Keyglove code, but it also involves a hardware modification of the Bluetooth module for the Keyglove Kit (which is a custom design). The Bluetooth module PCB is actually already sitting on my desk, but I need to get a new solder stencil so I can build the thing since there are quite a few surface-mount parts. I should be able to get the stencil within two weeks.
Second, I need to update a few things in my underlying code library for interfacing with motion sensors for improved speed and reliability. What I have now works most of the time, but it has no error recovery. When something doesn't communicate properly on the first try, everything locks up. This is obviously...well...bad.
Third, I need to work on that strain relief issue and experiment with a few different approaches. I'd like to get the hardware as done for the kits as soon as possible, since firmware updates are trivial for technically-inclined end users (e.g. many of you), and I'm sure those of you who pledged in return for kits would love to get your hands on them. This third step will also involve continuing to look for a company who can build the physical glove component with all of the sensors on it. I know there are people out there who can do this, but I just haven't connected with them yet.
That's enough for a short list of near-term to-do items. The new Bluegiga position is taking the majority of my time right now and for the next couple of weeks at least, but I'm determined not to relegate Keyglove to the bottom of the priority list. Keep your eyes open for more news, and thanks again four your support!