The world is in love with the "Internet of things" but we are using old tools to solve the problem. While we had no choice but to use native apps on our phones for this first generation of smart devices (e.g. Nest) it can't scale. If we believe in Moore's Law at all, we'll have hundreds if not thousands of these devices in our lives in a very short period of time. It just doesn't make sense to use apps as our primary interaction tool. The Physical Web is an approach to 'infuse' the web into physical objects so you can just walk up and use any device, on any platform, with just a single click.
The Physical Web
I’ve been a professional designer for
nearly 30 years and I’m more excited
about what we, as the design community,
can do more than ever before. It’s an
amazing time to be designer for the
simple reason that things are moving so
fast, the rules changing so quickly, that the
designer mindset has never been more
But that doesn’t mean it’ll be easy.
Let’s me tell you two stories.The ﬁrst
is with my Kindle, which has always
been able to synchronize your book
progress across all your devices. One
day I bought a best seller and it
offered me an audio book for a few
dollars more. What amazed me was
that when I went for a run, my audio
progress was also synchronized. I
literally did nothing and things worked
beautifully between all my devices.
Contrast this with the bluetooth headset
itself. I got up one night to go into the
kitchen to ﬁnd it on the kitchen counter.
Worried that it was on, I used the ‘mult-
function button’ to press-hold it off. It was
one of those devices that speak
commands and it chirped enthusiastically
‘Power on!” Of course, they automatically
powered down, there was no real need
for me to have ﬁddled with them at all.
Just as I was smiling benigning, it
continued “Connecting!”Which confused
me, my phone was in my bed room far
away but I was worried that it would light
up my phone and make noise, bothering
my wife. I hurriedly tried the press-hold
command again, but somehow got it
wrong it, causing it to chirp again
“Dialing!” <Two stone cold heart beats>
WHO are you dialing at 2am in the
morning?The last person I called was my
wife, who kept her phone next to her
bed. Let’s just say I am the unsung poster
child for the consequences of bad
How will we approach the Internet of things?
When we build products around the
internet of things, will it be a seamless
easy experience or a complex cascade
This isn’t an idle question because so
many people have a fairly naive approach
to the internet of things, a shallow
‘everything thing is awesome’ approach
but I feel fairly certain we’re heading for
a backlash, a UX disaster that if we’re
not careful is going to set us back for
Let’s step back a bit and start with an
innocuousTwitter conversation I had a
few months ago.This little animation was
posted and people were oohing and
ahhhing over how awesome it was. !
People were not only using words like
‘perfect’ but also ‘standard’ and ‘consistent’
Now don’t get me wrong, this animation
is lovely but the idea that we need to use
it forever or heaven forbid standardize
seemed a bit premature to me.
Thats a slippery argument. Are you really arguing that
we’ll use pull to refresh forever?
So I sent out this slightly provocative
tweet:Well, you’d think I shot Pope
Francis... The fanboi backlash was strong,
really pushing it as a standard. Not
because the design is ﬂawed, but if there
is one thing I’ve learned in this industry is
that things just change. These days, style
guides have a half life of about 12
months. So I pushed back a bit
The reply was that at the steering wheel
has been around forever! It’ll never
change, it’s perfect!You think the steering
wheel is a standard?
How adorableHow adorable
If you have any sense of history, you know
how silly it is to believe that the steering
wheel is a ‘standard’. Here are some
quick shots I grabbed from just the last
few years.The steering wheel is being
experimented with constantly. Besides,
self driving cars are just around the
corner. Of course, they are still a ways
out but when they come in, Steering
wheels be as standard as buggy whips
The History of the Steering Wheel
But even if you’re dubious that steering
wheels are going away, it’s history is far
more nuanced, and troubled, than you’d
think. Initially we had no steering wheels,
just horse drawn cars.The horses, for the
most part, steered the cart.There were
reins of course but they were meant
mostly for “whoa” and “go!” commands
with a few suggestions at forks in the
Boats were the ﬁrst vehicles that needed
active human steering.A quick refresher,
boats had the rudder in the back and
when you pushed it over, the sideways
force on the boat helped push it into a
When cars came along, they copied
boats, not horse drawn carts.The engine
was in the back, the back wheels turned,
and even tillers were used to steer them,
not wheels.What is fascinating to me is
that even when wheels were used on
boats, they mimic’ed the tiller mechanism
to to turn left, you had to turn the wheel
right.This was even true for cars.The
completely copied the mechanics of
Here isThomas Edison himself, in a car in
1901. Notice that the steering is now in
the front but he’s still using a tiller to
steer the car.
The steering wheel "has met with such general disfavor in this
country that levers are used almost exclusively”!
User Unfriendly, Joseph J. Corn
And this is the critical point. So often we
lament slow user uptake and how
companies can be too far ahead of the
market but in many cases users are pretty
damn smart.The technology just wasn’t
ready! Rack and Pinion steering was still
many years away. It took several technical
innovations in before wheels actually
became usable.Tillers really were the
better product to use at the time.
The Shape of Innovation
There is a shape to innovation and I think
the CHI community needs to encourage
this greater sense of history.We see this
pattern over and over.!
Whenever we get a new technology (like
cars) we tend to look the past for
inspiration.We model tomorrows
technology on yesterday’s tasks.We
eventually ﬁgure it out but it’s a bumpy
road to maturity.Then along comes a
new technology revolution,
The Shape of Innovation
and we start the process all over again.
Once we kill the steering wheel, what are
we going to drag into the future of self
driving cars? Dashboards? Bucket seats?
The car will reset onto a new bumpy
road of maturity loosing the steering
wheel and all of the artifact that came
But we have recent examples of this
same pattern.The DOS prompt copied
theTeletype machine which was matured
into the interactive character menus of
Lotus 123.The graphical user interface
then reset the market and we started
over with a new set of tools.
And the same thing for phones.The
original phones had a single menu, based
on the computer menu which was
matured in the ﬁnal Nokia Series 60
phones. It wasn’t until the iPhone came
along that this entire approach was
retired and replaced by atouch based
The Shape of Innovation
There are two critical psychological
points to this Shape of Innovation.The
ﬁrst is this ‘tiller moment’ when we
inappropriately recycle an old model. But
the second, even more powerful is when
we reach this maturity point.We’ve
worked so hard to get to this point that
we’re very reluctant to give it up!There’s
a type of intellectual gravity well that
forms, it just feels so right that the desire
to look past it just fades away.This is what
I saw with myTwitter friend invoking the
steering wheel, he was standardizing on it
as he was overly attached to it.
The Shape of Innovation
The problem is that we see this gravity
well in so many products.Windows
attempted to mature into a tablet
product more than 10 years before the
iPad. Microsoft was so attached to their
model that they just couldn’t let it go.
The book, The Innovator’s Dilemma,
discusses this reluctance of companies to
create breakthrough products. It focuses
on companies desire to focus on their
core (high end) customers too much,
missing the new opportunities. I don’t
dispute this bu I feel there is another
strong force, this intellectual gravity well
that holds so many companies back,
making them extra conservative.
Everyone wants innovation
but no one wants risk
I saw this so clearly when I was
creative director at frog design. Our
clients always wanted big innovative
ideas but after we would bring them
in and do our thing, they would so
often push back, letting most of the
ideas fall on the ﬂoor. Everyone
WANTS innovation but no one wants
risk… We often joked at frog that we
weren’t in the design business but the
design therapy business. Much of our
job was working with the entire
company team, ﬁnding the hidden
risks and concerns and getting them
to look past short term concerns.
Risk-averse people are not afraid of the future…!
They’re just overly attached to the past
Because risk is a very tricky human
reaction. Risk averse people are not really
afraid of the future, there just a bit overly
attached to the past.
OH MY GODOH MY GOD
He hasn’t talked about IOT yetHe hasn’t talked about IOT yet
I realize some of you may be wondering,
He hasn’t really talked much about the
Internet of things yet.What’s going on?!
But the internet of things changes
EVERYTHING, from our designs to our
products to our very business models
and if we don’t understand this shape of
innovation and how it affects how we
start (being familiar with the past) or how
we get stuck in an intellectual gravity well,
we’ll never really think about it properly.
This is especially important as the IoT has
been verging on the insane.Take a look at
these market projections that have come
out every 6 months over the last 3
But even the way we are talking about it
is absurd.A recent article in Wired had
clearly silly scenarios such as warning if a
girl friend is nearby. It’s so clear that
technology is driving the discussion and
not reasonable use cases.We’re getting
excited just because it’s possible.
There’s this basic belief that if everything
is connected and smart then we can
basically do anything and the Jetson’s style
projections start spewing forth.We’re
assuming old science ﬁction beliefs.
Besides, we can’t really trust the Jetsons
at face value, notice that tiller handle
The basic problem is that the Internet of
things is such a broad term, it can mean
so many things. But let’s just pick the two
that are getting the most press: Smart
Devices and Home Automation.
Smart devices are things like the Nest
Thermostat and the Quirky Egg Minder:
single function devices. Home automation
is more a cascade of information where
my calendar talks to my alarm clock
which wakes up my coffee machine and
Let’s start with Home Automation. Here
is an example from NinjaSphere, a
kickstarted campaign that raised well over
it’s original asking price.The idea is simple:
by tracking your location within the
home, you can automate the lights as you
move through it. I don’t want to make fun
of this product, it actually has some
potential, as long as you’re a single guy
This ‘everything is awesome’ vision is
indeed inspiring but it’s a bit fragile
The issue I have with this approach is that
it reduces your home lights to a switch
and a single light!
It replaces the switch with a ninjasphere.
The problem is that it then has the
responsibility to fully operate that switch.
Room is empty Simple motion sensor
Wife is sleeping Better sensors
Dog is sleeping Even better sensors
Wife watching TV,
looking for remote
WTF… Just read my mind, would you?
Bedroom Scenario However, the UX designer in me is far
too painfully aware that humans are
messy, illogical beasts and simplistic if/then
rules are going to create a backlash
against this technology. It isn’t until we
take the coordinated control of these IoT
devices seriously that we’ll start building
more nuanced and error tolerate
systems.They will certainly be simplistic at
ﬁrst but at least we’ll be on the right path.
We must create systems that expect us
to be human, not punish us for when we
The IoT is not a set of if/then rules
There are so many times you can say
“sorry honey, that wasn’t supposed to
This is such an obvious thing to say but
being human means being illogical.You
can automate 90% of daily life but you
can’t make the remain 10% more difﬁcult.
Moravec’s paradox: In the artiﬁcial
intelligence (AI) ﬁeld, Moravec
commented that computers can be good
at intellectual adult tasks but quite
horrible at simple perception tasks a small
child could do.This divided AI problems
into two broad camps: HardEasy and
EasyHard. HardEasy were problems like
Chess which people thought required
cumming so were hard to do. It turned
out that by throwing hardware at the
problem and doing a (fairly) simple
search, a computer could play chess quite
well. Language transition would be an
example of an EasyHard problem.They
thought it would be fairly simple to just
look up all the words needed and lay
them down, but it was clearly much
harder than that. !
I claim that home automation is an
EasyHard problem.I’m not a luddite, I
strongly feel this is an excellent direction
to go but we don’t give it the credit it
What is theirTechnologyTiller? The problem is that they are trying to
mature the light switch by replacing it
with sensors and rules, but that’s an
EasyHard problem.What should be
matured are the lights themselves as they
deﬁne the problem. Much like the
steering wheel, there are low level
technical things that are keeping us from
making this switch. Once we evolve our
lights to be smaller, more easily switched,
we can create a much more responsive
system, something that isn’t as risky as
‘turning on the lights’ would be today.
Bedroom ScenarioTake 2
Room empty? Someone there?
Turn onTurn on
If we had this more robust lighting
approach we could redo the bedroom
scenario. If there is no one in the room,
then go ahead, turn on the full lights as
you would have before. Nothing much is
changed here. But if there is someone,
then just turn on the safe ﬂoor lights so
you can navigate the room without
bothering someone. However, have the
humility to ask the user. Extend the light
switch to them by, for example, vibrating
their watch, letting them know the system
WANTED to turn on the lights but didn’t
think it was safe, extend the light switch
to the user so they can make the ﬁnal
We need to be building systems…!
that EXPECT us to be human,!
not punish us for when we are.
The change is fairly small, we just need to
treat home automation with the respect
it deserves and create systems that keep
us in charge at important moments.
Now let’s move on to Smart Devices.
We’re starting to see lots of smart devices
Is there aTechnologyTiller?
We’re starting to see lots of smart
devices, each with their own app.This is
perfectly ﬁne when we have a few
devices but what of the future? If we
believe in Moore’s law at all, there will be
hundreds, even thousands of smart
devices in our lives. How are we going to
work make that work? Is there a
technology tiller here that we are trying
to drag into the future?
Apps don’t scale
The problem is that apps just don’t scale.
Right now everyone is focused on home
control, e.g. making your washing machine
orTV smart. But the much bigger prize is
the public space with buses, trains, rental
cars, vending machines, parking and of
course, every shop you enter.All wanting
you to use their app. It just isn’t practical
that users would preinstall these apps.
Just inTime !
There is a wide range of devices from the
nest down to bus stops (which are just a
steel pole stuck in concrete)There is a
continuum of device from standalone
processor to a tagged object that points
to a web page. But, from a design point of
view, they are all the same: they want
your attention and you need to interact
with them.The problem is that we are still
using our old school paradigm of ‘native
apps’ to deal with them.While I might be
ﬁne with an app for my Nest, am i going
to download an app for ever store I enter,
every smart poster to see, or every
smart museum I enter? As we move to
single use experiences, apps become
Mobile apps must die
It’s why I wrote Mobile Apps must die.
People thought it was a rant about web
vs native apps but that wasn’t it at all.
Every time I talk about the IoT I get questions that show that
people really, deeply don’t understand what it is about. My
favorite example is the smart toaster, the derogatory poster
child of the IoT.When people say that “I don’t want apps on
my toaster” I want to shake them by their shoulders! “That’s
*your* old paradigm, not mine. It’s too easy to criticize a new
technology using old concepts. Smart devices are not about
apps!They are about 3 basic layers of functionality: Discovery,
Control, and Coordination!
!Discovery: Finding my devices nearby. Most companies would
kill for just this basic feature. Depending on how clever they
are with the URL it can span goofy marketing page (boring) to
SPIME like deep interaction with my device history.!
!Control:A small increment in cost lets me control the device.
This same URL model has moved us from web site directly to
Nest because now I’m talking to MY device.While prices are
still high, ﬁnding the right balance will be tricky but as the costs
fall, the choice will become trivial.This needs ‘another Apple’ to
take the chance because once it becomes clear it is possible,
EVERYONE will want to jump into the pool.!
!Coordination:This is the hard one as it involves so much
cooperation. I love the overall vision but it will take time for
companies to get on board with enough standards to make
this happen.Think about today. I can hardly get Mint.com to
access all of ﬁnancial records, it’s constantly breaking down.We
expect massive data and control settings to work across every
world wide manufacturer?!
!My point is that we need to start with the ﬁrst two: Discovery
and Control.They are very much within our reach and offer
It’s important to understand how small
we’re talking about. Imagine just having a
little bit of data like a phone number or a
small paragraph of text. Objects could
just give you that little bit of information
you need. But it can extend to simple
web pages, like a printer support page at
work. Or even further to simple
interaction so I can pay for a parking
meeting or rent a bicycle. By losing apps,
you can link small and solve a much wider
range of interesting use cases.
The Web needs a discovery service
2. Ranking 3. Interacting
This is NOT the internet of things, but a
building block that can lead up to it.There
are 3 areas to discuss: 1) Discovering the
public devices nearby 2) Gathering meta
data to help rank those devices and 3)
Letting the user choose and interact with
that web page.Again, this is similar to the
steering wheel in that the web is perfect
for this but we don’t have the right
technologies in place to unlock it.A
simple discovery service would unlock a
whole to range of both basic info and
The Physical web
Everything has ‘a web page’!
‘Instant interaction’, walk up and use!
No app install/manage/delete!
Encourages entirely new lightweight models
Bridge web & physical devices
The solution is “The Physical Web” a way
to bridge physical devices and the web.At
it’s core, its a simple means for devices,
like a zip car, to broadcast a URL so any
web enabled device can detect that URL
and use it. Everything gains a web page.
This unlocks the super power of the web
and makes instant interaction possible.
You completely remove the need to
manage apps and most importantly, it’s so
simple and light weight that it encourages
new riskers products that wouldn’t have
been considered before.
In a sense, I’m looking to create a “just in
time” ecosystem. where a range of
devices can broadcast a URL and range
of smart screens can be looking for them.
Now are these ideas utopian, even a bit
naive? Absolutely.This isn’t an easy way of
looking at the world, it’s full of stumbles
and dead ends. It’s not enough to be
open source, it has to be big and
audacious, just like the original internet
our time horizon for innovation has
become weeks not decades. People
forget that the delta between Netscape
and Gmail was 10 years. It takes time for
technologies and markets to move. I’m in
this for the long haul and willing to tackle
this problem in small bites.
Remember these guys?They were the
‘pre-web’, for a while, they were much
better than the web.They tried to be the
Apple of the web, a safe place where
things were well under control.The web,
at least initially, wasn’t as good, but it
eventually overran all of them for the
simple reason of scope and reach: they
just couldn’t compete with the
Look at this this way: FedEx couldn’t exist
without a municipal road system. Great
companies are built on great
infrastructure, usually public infrastructure.
It’s actually very simple, the world is really
only split into two groups: truck ideas and
The problem is that everyone wants to
build trucks. Building roads just doesn’t
seem very sexy.
Apple Samsung FacebookCisco
The other problem with truck ideas is
that they they tend to build their own
trucks often with their own roads! In
order to maximize proﬁt, they build an
ecosystem that locks out others.!
I’m not naive, I appreciate the business
world isVERY competitive and if you
don’t protect yourself, you’re vulnerable.
But it’s such a colossal waste of energy.
Maybe it’s such a dog eat dog world
BECAUSE everyone is trying to play king
of the mountain?
Does anyone know this guy?This is my
new hero. Malcom McLean was the ﬁrst
business man to see outside of this
defensive model of capitalism.This is the
guy that invented container shipping.
Truck Train Boat
Back in the 1950s you had to load and
unload all sorts of different sized cargo
from trucks to trains to boat. McLean
created a standard container size that let
you use cranes to do the unloading, it
was 36x cheaper than doing it by hand.
He made a mint
But you know what he did? He gave it all
way. He had patents on everything but
made them all royalty free.Why? For the
simple reason that he realized he was
working on road problem. If everyone
used his system, it would create a much
bigger pie.And guess what, he made even
more money.The guy cashed in big time.
What would you rather have?
75% of this 25% of that
It’s surprising that a businessman of the
1960s has so much to teach us today.
What would you rather have 75% of this
or 25% of that?
But the big companies won’t go for this!
I keep hearing that big companies won’t
go for open systems like this. Maybe not.
But I’m encouraged by the open
hardware and software movements.They
enable small companies to do so much so
easily. Combine that with the increasing
interest in crowd funding and there are
likely to be many interesting experiments
in the near future. If only of few of them
start to create an open ecosystem that
starts to gain traction, it will create PR
pressure of the bigger companies to
In thinking about the future, it’s easy to be
blinded the the giants of the day.The
iPhone is great, it was a major step
forward but to keep worshiping it,
copying it’s model is just getting stuck in
an intellectual gravity well.
The Shape of Innovation!
Apps are ourTechnologyTiller!
We need the Physical Web
The basic point isn’t that surprising, the
IoT changes everything. But we need to
be aware of the shape of innovation to
make sure we’re not getting stuck.Apps
are ourTechnologyTiller and holding us
back but implementing the Physical Web,
a new way for us to interact with the
Power to the people