Google I/O 2013 – Jean Wang – 7 Techmakers and a Microphone


FEMALE SPEAKER: So now we’re
going to shift, and we’re going to hear an amazing talk by
a hardware lead at Google. Jean Wang has a PhD
in electrical engineering in photonics. She worked in the optical
communications and then came to Google. She’s the hardware lead
on Google Glass. As a board member, I
have Google Glass. I love it. It’s so much more natural
than a phone. And now you get to hear about
how she did this. Thank you. [APPLAUSE] JEAN WANG: Hi, everybody. Smile. So I’m very happy to be
here with you today. I’d like to share with
you my Glass journey. It’s a three-part story that
spans across multiple domains. And while it is yet unfinished,
I hope you share my enthusiasm for what
is yet to come. Before diving in, I’d actually
like to float this number briefly and plant it
in your minds. It has significance to me and
the story, and in the meantime, please feel free
to question the number base that I’m using. So, Glass, like many ideas,
started as a blank canvas. As part of the initial team,
we came into the project largely inexperienced
in the field of head-mounted displays. Instead, we had the ability to
divide and conquer through our ability to learn and adapt. While some of us focused on
optics, like myself, others worked on cameras, sensor
systems, electronics. And so we also came in with the
fundamental belief that technology had ripened to the
point where we can make a significant impact and create
something of significant use. And that has led us into the
discovery phase, to find out just what it was that could be
done to basically minimize the barrier to information access,
and to make us connect with each other in terms of accessing
our social sphere. As part of the discovery
process, we looked at existing products in the mobile space
and wanted to really understand where it was that we
can make an impact in terms of form factor. For example, whether it was
bulky and had to fit into our pocket or carry around a bag
for it; ease of access, whether we had to pull it out
of our pocket or a bag to actually engage with it; and
also ease of use, whether we were blocked when we were using
the device, or whether we could look straight ahead and
talk with people and not be distracted. So to push on these pain points,
we did a number of technical experiments. Here’s a helmet-mounted display
that did not allow us to be hands free
or wires free. And in the bottom, you can see
here, is an early optical experiment to help us understand
the technical complexities in terms of how
optics sizes actually impact the performance and
the image quality. So with these considerations
in mind, we dived into selecting components that
we could use to make a viable solution. So in the vast array of options,
we looked at what it was we could just reuse
and customize. Or was there something that we
had to invent for ourselves? So we played around with a bunch
of pico projectors and projectors. We looked at optic components
such as lenses and prisms as well as interface devices– touch pads, microphones,
speakers. And underpinning the whole
system was electrical components and board design. So in this huge selection of
choices and huge array of options, how do we
down-select? And so we filtered using
two key criteria. One was power consumption,
and a second was weight. So these two are basically
factors in enabling us to first, wear a system all day,
and second, actually allow us to want to wear a
system all day. So we headed into the second
phase of prototyping. Here are six example generations
of glass prototypes. Starting at the very top left,
we took the lowest hanging fruit of simply putting mobile
phones to the side of the head, and a pico projector
shining directly into the eye to provide the image. And so this was a bit bulky,
as you can imagine. It came in at 167 grams. And we wisely decided that we
could lighten the load by taking the mechanicals
off the phone. And that’s our second proto
shown right up there in the middle top. And so as you can see, over
time, we shifted from using off-the-shelf components, such
as safety shields and frames, into 3D printing
our own frame. And that’s shown here
in the third. And you can also see that we
migrated into making our own optical design and architecture,
which is something we carry
through today. And as you gradually evolve to
the very end, you see that more and more we use customized
components. And we look at making our own
plastics and metals to get to where we are today, which is
essentially the essence of what Glass needs to be. This provides the core, and
provides a very minimalistic system, that allows us to run
around and be happy when we’re playing with the system. And in addition to prototyping
with the hardware, Glass derives its use from
easy to follow UI. And this is by definition
actually easy to make. Along the way, we experimented
with a number of different interaction concepts, for
example, head gestures. And that recent SNL skit was not
too far from our previous truth, if you’ve seen it. Also, we experimented with
porting the keyword functionality to the touchpad. But we found that was a bit too
complex, even when running in a predictive word mode. We also tuned the number
of lines of text– the font, the size– to basically optimize
for readability. So through all of these
things, we found that minimalizing and stripping down
to the very core allowed for the best use case
and experience. We also found that minor head
tilts and voice commands allowed us for hands-free
experience, which is very important when we’re
busy dancing or riding that roller coaster. On the flip side, having the
touchpad and the camera button allowed us for more quiet
modes of interaction. So with the hardware and
software solutions converging into a viable state, the team
actually turned to the question of production. It is possible to manually tweak
and baby one device, to get it to be close to ideal
in terms of cosmetics and functionality. However, we’d like to share
Glass more broadly. And so that, to us,
quickly turns into the question of scale. How do we build not 1, 1,000,
100,000, but upward towards a googol of devices? Layering in automation
is key to this. And that means that not only
do we have to build Glass itself, but we have to build infrastructure that builds Glass– this floor space, the
equipment, the whole kit and caboodle. And to add a bit more
complexity, we’d like to honor individuality and human
preferences. And that to us, has meant color
options, and that leads to color complexity. How do we make sure that on
the device itself, part to part, the colors match? And how do we make
sure that between devices the colors match? So production requires the
solutions to all of these questions and beyond. Through these three phases of
discovery, prototyping, and production, I have found that
Glass is both an innovation and a reinvention. By integrating highly compact
and new components, we’ve been able to extend far beyond nickel
head-mounted displays, beyond the last couple decades
of individualized movie players, game consoles,
and aircraft systems. On other hand, I think that
Glass is a way in which we can examine and develop new fields
that we can push beyond and learn to be better connected to
each other as well as share perspective and have access
to information. And all of this will allow us to
better empathize as people and sympathize. So coming back to that number
at the very beginning. It’s been exactly, as
of today, 1,111 days since I joined. And this is a picture I
would not be able to take without Glass. I am paragliding over Coronet
Peak, New Zealand. And just by looking at it, I can
recall what it was like to soar with the birds
and to take in the beauty of the landscape. I have often been asked what the
a-ha moments have been in the project thus far. And while I have a few, I
am more excited about what is yet to come. There is a first tier, for
example, a bar code, QR code scans, where we can get
just-in-time reviews of movies, books, that menu trace
we’re just about to take, but we’re not quite sure yet. Then there’s also never getting
lost and not bumping into things while you’re
looking at directions. And then there’s a whole
other level. Recently one of my colleagues
exchanged his wedding vows while wearing the device. And I get goose bumps when I
imagine him reliving that when he is old and gray and seeing
the reflection of a future from his wife’s eyes. And then there’s also a
physics instructor who recently toured CERN
with the device. And he did a Hangout back to
his classroom in Michigan. And he was riding his bike down
the tunnel of the Large Hadron Collider showing his
classroom just how cool it was, while taking live questions
on the spot. Then there’s also another
example where Glass can mean life or death, that it can
help you translate to a medical doctor in a foreign
country that you’re allergic to penicillin. So through this process, I think
that there’s just so much potential in what is yet
to come, that in pushing towards a new perspective on
information sharing and perspective sharing, we’ve come
upon a new environment Glass can engender. And when I have time to pause
and think about how far we’ve come, I feel privileged to be
part of such a dynamic team. Together we’re striving forth
to bring forward this new perspective. And I count you, as developers,
as part of this growing group. There’s just so much potential
and positive impact in keeping in the moment, where information
is just at your fingertips and knowledge is
a twinkle in your eye. And with that I look forward
to your aha moments, and we are back to discovery. [APPLAUSE] JEAN WANG: Thanks everyone.

Leave a Reply