EINSTEIN &THE IOT:
W. David Stephenson
March 1, 2017
Good morning! I’m going to talk about the Internet of Things, or IoT, today, but my point is actually far broader.
I think that point is vital if you in this room are to achieve your potential as the business innovators of the future. It concerns how we use innovative new technologies —
and, most important, how we think about them!
“YOU CAN NEVER SOLVE A PROBLEM ON
THE LEVEL ON WHICH IT WAS CREATED.”
- ALBERT EINSTEIN
The reason for Einstein’s name in my title is that he said, “You can never solve a problem on the level on which it was created,” or as it’s usually paraphrased, “you can’t solve a problem with the same thinking that created it.”
That reminds me of the executives — and there were many of them — in the 1990s who didn’t actually use email themselves, but dictated them to their secretaries and then had the emails and replies printed out and stacked
neatly on their desks. Sure, they were transmitting these messages more rapidly, via a cool new technology, but they were using them just like traditional printed memoranda. They totally missed out on email’s transformative potential.
Hmm. Wonder how they thought about embedded hypertext? Probably thought the lines were important because they were underlined!
Similarly, we simply won’t realize all of the IoT’s beneﬁts if we shoehorn this technology into our conventional ways of doing business. We must make dramatic changes in the thinking and assumptions that underlie our current
economy and, unless that old mentality is abandoned, will create a drag on the IoT.
That’s where you come in, because millennials aren’t handicapped by past practices. We’ve already seen, especially in your embrace of the sharing economy, that you’re willing to part with long-held assumptions, and that
you leave conventional thinkers scratching their heads.
Now don’t get me wrong: I get excited as the next guy at the incredibly cool new devices and services made possible by the IoT — particularly love this one, the
AliveCor. It attaches to the back of your phone, and, in 30 seconds, can give you an FDA-approved accurate ECG reading of your heart
— BUT, we’re falling way short of the IoT’s potential for change if we stop with cool things, because I believe..
THE IOT CAN
COMPANIES & ECONOMY
.. that the IoT can fundamentally transform our companies and economy. Note the can — it won’t be automatic.
“For companies grappling with the transition (to
the IoT), organizational issues are now center
stage — and there is no playbook.We are just
beginning the process of rewriting the organization
chart that has been in place for decades”
—Heppelmann & Porter
What got me thinking about this potential for more radical change was a quote from Heppelmann & Porter’s second HBR article about the IoT, focusing on management issues. They said, “For companies grappling with the
transition (to the IoT), organizational issues are now center stage — and there is no playbook. We are just beginning the process of rewriting the organization chart that has been in place for decades”
I’m a very right-brained intuitive thinker, and what that triggered in my mind was something I’d been talking about for while regarding companies’ problems before the IoT, and the dramatic solution it oﬀered to these problems.
What if there was a condition that aﬀected every single human — no exceptions — that meant it was impossible for us to see inside things and determine their real-time status.
This would be an example of what business theorists call “wicked problems” – thorny ones for which not only the solution is unknown but also the existence of the problem itself, so the problem is only understood when it is
solved, not before.
In this case, that was because we were unable to see within much of the material world, from assembly-line machinery to how products were used once they left the factory. Collective Blindness kept us from really knowing a
wide range of critical facts about our businesses and how to manage them, from when a critical assembly line part might break to how our customers actually use our products and how they work in the ﬁeld.
As a result, we unconsciously lowered our sights, and adapted to these limits in knowledge. We improvised workarounds to operate as best we could with insuﬃcient information. They worked OK, but they did lead to higher
repair costs and liability, poor decision making, assembly line ineﬃciency, and unsatisﬁed customers, among other eﬀects.
Collective blindness also aﬀected management.
Typically, in a factory like this, trying to determine equipment’s status meant an individual worker had to observe and record a gauge’s reading on paper — usually
on a regular schedule. But what happened if there was a pressure spike in between the readings? What if his supervisor forget to check the form? What if there was a
critical factor, such as metal fatigue, that couldn’t be detected by a gauge?
source: Dynamic Ofﬁce Supplies
It was equally diﬃcult to share what data you could gather: typically the information was paper-based, and had to be distributed from a central location to those
who it was thought might need it.
source: wikimedia commons
Is it any wonder that these restrictions led naturally to the linear and hierarchical organizational structures that still characterize organizations in the Internet Age,
and that it was frequently senior managers who decided who got access to information and when?
Even if the Internet and new communications technologies have allowed us to remove some of those limits to communication, and even though there have been
some successful eﬀorts to “ﬂatten” hierarchies and remove layers of management, the ﬂow of data in most companies remains downward and lateral.
NOW, EVERYONE WHO
NEEDS REAL-TIME DATA CAN
SHARE IT INSTANTLY!
Here’s where I think the truly most profound and potentially far-reaching aspect of the IoT manifests itself: for the ﬁrst time, everyone who needs real-time data to do their jobs more eﬃciently and/or make better decisions,
can share that data — instantly.
Instead of the traditional “need-to-know” criterion for allowing access to information, we must make a radical transformation. Going forward, the default assumption must be the polar opposite: that management should have
to justify limiting access to data.
What this sharing of real-time data does more than anything is to provide “ground truth” to everyone involved: i.e., instead of having to act on inference, as in the past, when it was impossible to peer “inside” things, now
decisions and actions can be based on information provided by continuous, direct observation (albeit by sensors) instead. Decisions and strategy will be fact based.
This shared, real-time data about things and their status will allow an end to the old hierarchical and linear processes that I mentioned, which have been a given of
management theory since the beginning of the Industrial Age.
Instead, we can move toward circular companies, without data silos isolating individual departments or even external functions such as supply chains, distribution
networks or maintenance.
This graphic will give you the idea: a seamless circle revolving around a hub of real-time, shared data.
• Increased efﬁciency. Functions integrated.
• Improved morale.Workers feel valued, informed.
• Increased creativity. Different departments &
individuals collaborating discover things none could
on their own.
The results will be incredible when everything is linked.
• Eﬃciency will increase because various functions will be integrated.
• Morale will increase because workers will feel valued and informed.
• Perhaps most important, creativity will increase. For the ﬁrst time, diﬀerent departments and individuals will be able to collaborate in real time, and the resulting
dialectic will mean they will discover things none could on their own.
But we’re a long way from achieving the vision of the Circular Company, and we’re still very ﬁrmly rooted in two hundred years of hierarchy and linear processes.
So how do we get from here ….
That’s where Dr, Einstein comes in. We won’t be able to make the transition unless we abandon deeply ingrained attitudes from the past.
I call these new attitudes ….
4 ESSENTIALTRUTHS OF IOT
• Make Privacy and Security Highest Priority
• Share Data
• Close the Loop
• Rethink Products
The Essential Truths of the IoT:
• Make Privacy and Security the Highest Priority.
• Share Data.
• Close the Loop.
• Rethink Products.
I suspect it will be you millennials who will help us make this transition.
Let’s explore each of them and why they are so important to realizing the IoT’s full potential.
#1: MAKE PRIVACY AND
SECURITY HIGHEST PRIORITY
I used to list this as the fourth IoT Essential Truth to consider at all times, but I’ve decided to leapfrog it to No. 1, because the sad reality is that not enough companies take privacy and security seriously — right from the
It’s critical to Truth #2, Sharing data, because sharing data in new ways immediately creates risks that it will be shared inappropriately, especially if it concerns highly personal issues such as health and ﬁnances.
Newsweek, Jan. 29, 2016
Recently there has been a spate of scare headlines about Shodan, which bills itself as the “search engine for the Internet of Things.” I ﬁrst warned three years ago about the need to take strong security measures, such as
NOT using the Real Time Streaming Protocol (RTSP) without requiring password authentication, in order to avoid having a data feed from your device showing up for anyone to monitor on Shodan. Yet, the amazing array of feeds
Shodan provides today, from baby cams to marijuana plantations, shows that privacy and security are still not a high enough priority for many IoT developers.
Wired, July 21, 2015
This is an issue that I know more about than most in the IoT world do. You see, in a prior incarnation, I was was a crisis management consulting for Fortune 100 companies that had done dumb things, and now were in
serious, serious trouble with the public, government, and media.
Most frequently, the decision makers involved were engineers, and they seemed particularly insensitive to the role that FEAR played in the public’s reactions to their deeds. A common comment by the engineers, understandable
because they were so left-brained and analytical, was to dismiss the validity of fears because they weren’t fact based and rational. What these engineers didn’t seem to understand was that, no matter what its basis, this fear was very
real to the public, and they dismissed that perception at their risk.
That fear factor is even greater with the Internet of Things, because it often involves personal data mentioned above, or products that, in the wrong hands, could actually kill someone, as with the Jeep that two “white-hat” hackers
took over by exploiting a security ﬂaw not in the car’s drive electronics, but its lowly entertainment system.
Most dramatically of all, in a disaster that caught everyone’s attention, in late October the hackers behind the Mirai malware unleashed it by attacking a huge
number of IoT devices whose manufacturers either provided no protection or only rudimentary ones (speciﬁcally it attacked devices that used one of more than 60
common factory default usernames and passwords — such as “password” — doh) , infecting them with the Mirai malware. Then it launched massive DDoS attacks on
New Hampshire, temporarily bringing down sites such as Twitter, Netﬂix and GitHub. Got your attention about the need for IoT security??
CAN’T ADDTHEM LATER
That’s why a ﬂip remark by a young developer at the Wearables + Things conference who blithely dismissed the fact he hadn’t addressed security yet “because
we’re just a startup – we’ll get to that later” was so distressing to me: with the IoT, you don’t have the luxury of waiting until later to add security and privacy features.
They must be an integral part of the product or service from Day One.
EASYTO LOSE IT
In part that’s because of another reality of crisis management: it is very hard to build public conﬁdence, especially in a new technology where people have
immediate fears of personal information being shared intentionally or by hackers – and incredibly easy to lose it.
Equally important, fear doesn’t make nice distinctions between the good guys who design in security and privacy protections and those who don’t: if there are a
few high-proﬁle security leaks such as Mirai, it could undermine the entire industry.
PRIVACY BY DESIGN
Privacy and security are likely to be never-ending challenges to the IoT industry, because of the proliferation of devices, hackers’ skills, and the vulnerability of sensors, the weakest link, which must be
designed to be cheap and to last for many years, making them more easily hacked as hackers’ skills increase.
I’d advise that you follow the lead of some companies that adopt a “privacy by design” strategy: As Gulio Corragio puts it “the principle of data protection by design requires data protection to be
embedded within the entire life cycle of the technology, from the very early design stage, right through to its ultimate deployment, use and ﬁnal disposal. This should also include the responsibility for the
products and services used by the controller or processor.” Unlike the young entrepreneur who made me cringe, privacy can’t be bolted on after the fact.
A key tool that can bring about this industrial-grade IoT security is blockchain, the underlying technology for the bitcoin cyber-currency. It’s particularly relevant to our discussion, because blockchain makes
absolutely no sense in the old paradigm where the way to keep data secure was to hoard it. With blockchain, it is secure speciﬁcally because the data is chopped into segments — blocks — each identiﬁed
by a 32-digit hash and stored on a single computer, then many, many other blocks are attached to it, becoming a blockchain. It is secure because, once entered the data can only be changed by consensus
of all those who hold the individual blocks, making it easy to identify potential attacks.
To add a new transaction to the blockchain, all the members must validate it by applying an algorithm to conﬁrm its validity.
Individual eﬀorts aren’t enough: more industry-wide collaborations, such as the IoT Trustworthy Working Group are critical. Equally important, though controversial in some industry circles, IoT companies should become
active participants in collaborative eﬀorts such as the FTC’s workshop in 2014 to create tough but workable IoT government regulations. These regulations, provided they are performance-based, rather than prescriptive (which would
limit the development of breakthrough IoT technologies), are important to building IoT credibility in the public eye: there have to be penalties for those who don’t take privacy and security seriously because the alternative will be public
#2: SHARE DATA
After assuring privacy and security, sharing data is the next most important Essential Truth, because it is so diﬀerent from the principle that has been ingrained in us for so long: hoard data.
One day in 1789, a young man with few possessions sailed from England to the New World. Few possessions in his hands, that is. In his head, he held the plans for Arkwright's famous woolen mill. It
was illegal to take factory plans from the country at the time because they gave England such an economic advantage, but no one knew what was stored in young Samuel Slater's memory, and when he
arrived in Pawtucket, Rhode Island he used that knowledge to build America's ﬁrst woolen mill, becoming the father of America’s industrial revolution.
For the longest time, the kind of proprietary knowledge possessed by the British mill owners was the way to proﬁt: in a zero-sum game, if you had knowledge that I didn't, you were a winner and I was
a loser. As recently as the 1980's, the so-called "Massachusetts Miracle" took place because seemingly smart mini-computer companies built their success around proprietary operating systems that made
their customers dependent on them — that is, until open systems came along.
Perhaps the most obvious beneﬁt of sharing IoT data is in the ﬁeld of consumer products. Phillips, the Dutch electronics company, was among the ﬁrst to realize this. In 2013 they released the Application Programming
Interface (API), for their Hue lights. Although individuals had already reverse engineered the interface to create hacks, according to Hue System Architect George Yanni, “.. we actually want to help and grow and encourage this
community, and give them tools and proper documentation. Also, we want to give them commitment that this is the API and we’re going to support it and it won’t change overnight.” As a result of this conscious decision by Phillips
to give up control, developers have created a wide range of apps to give the Hue lights new versatility (want to make a “grand entrance” [ whatever that might be!] with your Hue lights? There ‘s a IFTTT – If This Then That – “recipe” for
that!). Even more impressive, when multiple companies are willing to share their APIs, developers create apps that can trigger simultaneous actions by multiple devices, even ones from diﬀerent manufacturers. Now, using Apple’s
HomeKit, you can say “hey, Siri, it’s time for bed,” and the Hue lights will gradually dim, the Schlage front door lock will lock the door, and the EcoBee thermostat will lower several degrees.
Because data is shared, it becomes more valuable to each participating company, something that was inconceivable in the old days of hoarding data to gain an advantage.
The most important result of that kind of willingness to share data is that it triggers “network eﬀects,” the term coined by Ethernet inventor Robert Metcalfe to describe how each individual device become more powerful as
more and more of them are networked. In this case, the data becomes more valuable by being shared, as do the products that are activated by that data.
Think what a profound transition that is from the days of proprietary information conﬁrming a strategic advantage: now companies such as Phillips that encourage sharing of data gain a competitive advantage speciﬁcally because
their products interface seamlessly with those from other companies, making each of them more valuable when combined than they are in isolation.
Sharing data is also a key contribution of the IoT to corporate organization and eﬃciency in ways that were also impossible before.
Think what a transformation is possible now that various departments whose work necessarily also eﬀects and/or depends on several other ones (for example, the product design department and the marketing department)
can now simultaneously study the implications of real-time data about how products are actually used in the ﬁeld (provided, that is, that management will allow them simultaneous access!) whereas that was a total mystery in the past,
and whatever information was available was probably passed along laterally and sequentially.
The potential to examine the same data simultaneously, and, even better, make collaborative decisions can lead to fundamentally better decisions and more eﬃcient operations, because each department brings diﬀerent
interests, needs, skills, and insights. When these are exchanged, each team gains a better understanding of the other’s priorities and concerns, and there’s a high likelihood that synergistic solutions will emerge from their dialogue that
no individual team would have devised working in isolation. That’s why I believe the circular company will eventually emerge from the IoT.
Combining insights from various departments and various skill sets can lead to ﬁndings that none of the groups could have found by themselves, due to the dialectic resulting from simultaneous access to the data.
When doctors in the Neonatal Intensive Care Unit at Toronto’s Hospital for Sick Children teamed with data scientists from IBM in “Project Artemis,” to examine the 90 million data points generated daily by the sensors on eight
babies’ incubators, they found something astonishing: a full day before there were any external signs that the infants had developed a serious blood infection (late-onset neonate sepsis), the team was able to diagnose the condition
from patterns in the data: they found a consistent pattern of decreased heart-rate variability and were able to administer antibiotics well before there were visible signs of infection, improving the babies’ chance of survival. The
doctors credited their interplay with the data scientists, who obviously brought a diﬀerent skill set to the issue (as well as the critical infrastructure to process and store the vast amounts of data), around the common set of data as
critical to the discovery.
Even more dramatic is what can happen when instantaneous sharing of data extends beyond the company walls. For example, SAP has designed a prototype vending machine which contains a variety of sensors,
documenting everything from how many bottles remain in the machine to – if the customer opts in – his or her normal product choices, which might result in a special real-time discount (not to mention the customer being welcomed
by name). The marketing beneﬁts alone are remarkable, but the machine also shares the real-time inventory data with the suppliers, so that a delivery truck that might be headed for a given machine might be instantly rerouted (on a
Machine-to-Machine, or M2M, basis) to another one where the inventory is lower because of a a spike in demand due to hot weather.
AUTOMATED SUPPLY CHAIN
Industry Week, Feb. 14, 2013
Similarly, applying the principle of real-time data sharing to a factory setting, assembly-line data might also trigger M2M reordering of components from a supplier
(which, I believe, might result in “re-shoring” of industrial jobs, since the beneﬁts of true “just-in-time” deliveries would outweigh any savings from a supplier half-a-world
away whose delivery time would be in weeks, not hours).
#3: CLOSETHE LOOP!
Closely allied to the second Essential Truth, the third requires that we must get rid of the old linear ﬂow of information, and instead substitute circular, closed-loop processes in which the data ﬂow leads to analysis, which in
turn leads to ﬁne-tuning of the process or other feedback that allows for continuous improvement.
Hence, the circular company.
For example, the diﬃculty in the past of gathering information about how cars were actually driven meant that auto manufacturers operated in a vacuum about what improvements were needed. Even worse, the diﬃculty of
gathering information about the cars’ performance meant that customer opinion, a potentially important source of data, was skewed: it was so diﬃcult to submit criticisms and suggestions that those who were either wildly in love with
the car or who hated it were willing to make the eﬀort required, leaving out the opinions of the vast majority of users who weren’t at either extreme. By real-time ﬂow of data from its cars (“iPhones on wheels, as some company wags
put it because of their extensive instrumentation), Tesla can get a much better picture of actual driver experience.
According to GE’s Vice President and Global Technology Director William Ruh, the feedback loop from the company’s many sensor-laden products, from jet turbines to medical imaging devices , has reduced the time it takes
for them to design upgrades: “G.E. is adopting practices like releasing stripped-down products quickly, monitoring usage and rapidly changing designs depending on how things are used by customers. These approaches follow the
‘lean start-up’ style at many software-intensive Internet companies. We’re getting these oﬀerings done in three, six, nine months,’ (Ruh) said. ‘It used to take three years.’
Now GE and others are using digital twins — copies of devices in the ﬁeld that are displayed on terminals in the labs, complete with real-time operating data, to see inside these devices and be able to pinpoint problems or
opportunities to upgrade them.
One major beneﬁt of the circular data ﬂow is that it increases accountability.
In the past, a department that was involved in the early phases of a project often didn’t receive objective feedback on how the project actually unfolded, minimizing their responsibility for the outcome. One story, apocryphal or not,
summarized the problems with this linear data ﬂow: reportedly GM found out only the ﬁrst time an owner of a new model in the 1970s went in to get an oil change for the ﬁrst time that just changing the oil ﬁlter required dropping the
entire engine, a costly and time-consuming process.
Why? Because no one thought to involve a shop mechanic in the design process, and no one else ﬂagged the problem, because each step in the design process took place in isolation from the others. Now, with the IoT,
everyone involved with a project is literally “in the loop,” and both has a role in deciding the course of action and has instant access to the data about how it actually operates, so they see the consequences and can also be held
source: Stetson University
The circular ﬂow of data also allows various departments that previously worked in isolation and sequentially to collaborate and try a new approach of inter-departmental teamwork. Perhaps the best example of this in
current business is W.L. Gore, and its famous “lattice” management style — which I see as a step toward achieving the circular company vision. According to CEO Terri Kelly, “Some of the most impactful decisions at Gore are made
by small teams. Within any team you’ll ﬁnd people with very diﬀerent perspectives; they don’t all think alike—and we encourage this. We encourage teams to take a lot of time to come together, to build trust, to build relationships,
because we know that if you throw them in a room and they don’t have a foundation of trust, it will be chaotic, it will be political, and people will feel as if they’re being personally attacked. We invest a lot in making our teams eﬀective,
so when they have those great debates—where a scientist doesn’t agree with a sales associate, or manufacturing doesn’t agree with a product specialist—the debate happens in an environment where everyone is looking for a better
solution, versus ‘you win, I lose.’”
The circular data ﬂow, as mention previously, also means that everyone gets valuable feedback on the consequences of their actions and on how things actually work.
That leads us to the 4th IoT Essential Truth, that we must re-think products.
#4: RETHINK PRODUCTS
Products gain new stature with the Internet of Things.
This is largely because of the fact that the ability to monitor them constantly after they leave the factory means IoT products mean we’re no longer clueless about
how products actually work after they leave the factory ﬂoor, and remain a dynamic part of the daily operations, rather than an end product that’s no longer part of the
Products become players.
Perhaps the most signiﬁcant example of this transition is that IoT-enabled products improve product reliability, both for the end user and the manufacturer. We can now do “predictive maintenance,” intervening at the ﬁrst
signs of metal fatigue, decreased performance or other metrics. Instead of ﬁnding out only after the product has failed, sometimes catastrophically, a jet turbine can be ﬁxed the next time the plane lands, minimizing cost and
What you see here is an augmented reality tool, Vuforia, from PTC, which deﬁnitely puts that Collective Blindness I mentioned earlier to bed. It uses Augmented Reality so that the repair person can peer inside the device and
see an exploded view, coupled with real-time operating data. Since this can be done from another location, it means the mechanic can often tell exactly what parts need to be replaced even before getting out her tools, so that she can
bring the repair parts with her rather than having to send back to the warehouse for them.
In some cases, products redesigned around the IoT can actually be repaired, and even upgraded, remotely, because of two-way communication. In early 2014, GM and Tesla both faced product recalls. The GM one was
terribly handled, costing people their jobs, millions of dollars in service work, and, of course the lives of the poor car owners who were killed by the faulty ignition systems that triggered it. Tesla’s situation never escalated to the recall
level: there was a problem with the suspension that could under some circumstances cause a ﬁre. However, Tesla responded simply by sending out a software patch that was automatically installed in every Tesla one night while the
owners slept. Instead of having to oﬀer ﬁnancial bribes to get owners to come in for the necessary repairs as is often the case with a recall, 100% of the Teslas were ﬁxed.
CUSTOMERS PLAY ROLE IN DESIGN
In some cases, customers actually play the ﬁnal role in an IoT product’s design: car companies allow drivers to make push-button changes in the performance
characteristics – one engine can be used in various ways. Or, while it may be a less-than-earth-shattering development, women buying heels from Ishuu, a Lithuanian
start-up, can instantly vary their appearance by choosing a diﬀerent pattern on the app that controls an e-Ink insert on the side of the shoes.
PRODUCTS BECOME SERVICES
Similarly, this shift to being able to constantly monitor the product’s status can also allow a signiﬁcant change in pricing and marketing. No industry has been aﬀected as much by this shift as jet turbines. Rolls-Royce’s engine
sales have declined signiﬁcantly in recent years, but they’re not complaining, because they are instead “selling thrust” under the CorporateCare program– charging the customer based on the amount of time the engine is actually in
the air and working.
As part of the new deal, Rolls also monitors the engine’s operation on a real-time basis to allow for the predictive maintenance, and the airline can also, for an additional fee, receive the real-time data so that it can mash it up
with other real-time data: atmospheric data, fuel consumption and other factors to maximize operating eﬃciency.
• Make Privacy and Security Highest Priority
• Share Data
• Close the Loop
• Rethink Products
• Make Privacy and Security Highest Priority
• Share Data
• Close the Loop
• Rethink Products
As important as advances in technology and analysis are to the IoT’s future, it is these fundamental shifts in attitudes and strategy — the IoT’s Essential Truths — that will really make the diﬀerence in whether the Internet of
Things’ full potential is ever achieved. They are diﬃcult to make, because they are so diﬀerent from prevailing wisdom — remember what Einstein said — but will be more than amply rewarded by the opportunities they create. We’re
counting on you, because you aren’t mired in the legacy of old ways of doing business.
W. David Stephenson
Thank you, and now I’d like to take questions!