The Internet of Things (IoT) is by no means a new concept. The term Internet of Things was first coined in 1999 by Kevin Ashton, a UK tech entrepreneur, and the basic concept was first demonstrated in the early 1980s. The broad idea of IoT has also been peddled under a variety of names: ubiquitous computing, machine-to-machine (M2M), the Internet of Everything, ambient intelligence, and a few more besides.
While IoT has been around for a while, the 2020s will be the decade of widespread IoT adoption. Spending on IoT technology in the next decade will dwarf all previous deployments. IoT will reshape every industry on the planet. It will streamline business processes, create new value, improve customer service, make infrastructure more responsive, reduce maintenance costs, shift business models, and speed process automation.
IoT is set to transform almost every business, every city, and every home in the developed world. Ultimately, it will transform the lives of everyone on the planet. All of the main ingredients needed to make the Internet of Things viable are now in place: low-cost sensors, cheap computing, widely available connectivity, cloud storage, low-cost compact batteries, energy harvesting technology, and standards to connect all the pieces together.
The idea behind the Internet of Things is that everything—objects, infrastructure, machinery, appliances, devices, perhaps ultimately our bodies—will be infused with sensors and connected to digital intelligence. Connected cameras, connected cars, smart thermostats, industrial equipment, smart traffic signals, wearables, connected jet engines, and connected toothbrushes are all examples of the Internet of Things.
The Internet of Things builds a bridge between the physical and the digital worlds. Information flows back and forth across that bridge, from atom to bits, and from bits to atoms. Sensors allow the digital world to understand what is happening in the physical world and actuators enable the digital world to act within the physical world. For example, a moisture sensor allows an algorithm to detect when the soil is dry, and a connected sprinkler (an actuator) enables the algorithm to turn on the water. These connections between the physical and the digital worlds enable us to create smart, interactive objects, to build responsive infrastructure that optimizes the use of precious resources, and to augment and semi-automate business operations.
Decades of massive capital investment by semiconductor companies has paved the way for the production of tiny, cheap silicon chips that will populate the Internet of Things. Today, you can buy a computer for five dollars. The Raspberry Pi Zero has similar computing performance to a laptop from the mid-1990s that cost thousands of dollars. In 2018, IBM showed experimental chips they claim have the equivalent compute performance of Intel 486 microprocessors that powered $3,000 PCs in the late 1980s and early 1990s. These tiny slivers of silicon are smaller than a grain of salt, and IBM claims they will cost less than 10 cents to make in volume production.
Connected computing intelligence is now a cheap ingredient that we can stir into the recipe for every product and every piece of infrastructure. Intelligence and connectivity can be integrated at near-zero marginal cost.
Billions of sensors are sold inside smartphones each year. The average smartphone bristles with a multitude of sensors: accelerometers and a gyroscope to track motion, GPS to track location, a proximity sensor, a microphone, a magnetometer (compass), and two or more cameras. Many smartphones also include air pressure sensors, temperature and humidity sensors, heart rate sensors, gesture sensors, fingerprint sensors, facial recognition sensors, and more. The volume economics of the smartphone sector have driven down the prices of all these sensors.
Trillions of cheap, tiny sensors will be deployed in the coming decade. Steve Whalley is the former chief strategy officer of, and now a strategic advisor to, the SEMI-MEMS and Sensors Industry Group, an industry body that advises companies on the future of sensors. He is also an old friend. When I asked Steve about his expectations for future sensor demand, he told me, “Sensors will experience rapid growth in the coming decade as they provide the data that feeds the AIs that provide real insights and value in products. The deployment of 5G networks to enable greater bandwidth, speed, and lower latencies for new products will unleash an explosion of traditional and new sensors in the next five years and usher in the era of ‘Smart Everything.’”
Enterprise software company SAP estimates that by 2030 the world will be filled with 100 trillion sensors. That's more than 12,000 sensors for each person on earth.
Sensors enable us to make the world “programmable.” By bridging the digital and the physical, they allow us to write programs that can respond when things happen in the physical world; programs that manage precious resources and allow the world to operate more smoothly and efficiently. By sensing human activity, sensors enable us to make the world more responsive to human needs and wants.
Cheap, abundant computing and sensors will be used to build smart, connected products and responsive infrastructure. Wearable technology will give people superpowers. Analytics, fed by data from trillions of sensors, will improve business decision-making and boost operational efficiency. Internet of Things technology will lead to an explosion of product and business model innovation.
Sensors are an essential component of any future business innovation strategy. You can't manage what you don't measure. Sensors allow us to take action, in near real time, when and where we sense it is needed.
In the future, we will use vast networks of sensors to measure, manage, and improve almost every aspect of human activity. Sensors will provide insight into everything from local air pollution levels, to vineyard operations, to the mechanical stress on bridges. By understanding more about the world, we can intervene where needed to fix problems and improve the way things work.
Sensors measure the physical world so we can make decisions that are informed by data. For example, rugby and American football coaches use accelerometers embedded in mouth guards to understand the forces involved in impacts. Real-time data from mouth guards allows coaches and medics to assess the risk of concussion and take appropriate action. Over time, the data enables researchers to improve their understanding of the correlation between impacts and concussion.
Some sensors are dumb. Some are smart. Dumb sensors include accelerometers, temperature sensors, and gyroscopes. These sensors measure some aspect of the physical world and turn it into numbers that a computer can understand. For example, a GPS sensor outputs two numbers that describe longitude and latitude. A microphone outputs a stream of numbers that represent the waveforms of sounds. Smart sensors detect more complex events and information than dumb sensors. A smart sensor uses built-in intelligence to interpret the information it gathers. For example, a smart camera is a combination of a traditional digital camera and image-recognition software that interprets the images it captures. Retailers use smart cameras to measure the foot traffic a store receives, and how it varies over time. Cities use smart cameras to count cars, trucks, and vans in traffic surveys, a boring, laborious effort previously performed by people. Intel has demonstrated a smart camera that uses artificial intelligence to spot abandoned luggage at train stations. It recognizes and tracks both people and luggage and establishes relationships between them. If a person walks away from their luggage for more than a few seconds, the smart camera triggers a security alert. Smart microphones, deployed in a forest, can listen for the signature sound of a chainsaw and text alerts (and GPS coordinates) to a forest ranger looking for illegal logging activity.
In the 2020s, many of the objects in our lives will become far more intelligent than they are today. Smart, connected objects will make economic sense and provide important benefits to both consumers and manufacturers.
When we make an object smart and connected, it becomes a “service portal,” a device through which digital value and services can flow. When you embed intelligence inside an object, it evolves from being a simple “dumb” product and gains the potential to become a service offering. New business models can be built around these smart, connected objects.
A simple illustration of this concept is the notion of a smart, connected teddy bear. Many kids grow up with a teddy bear, or something similar. You probably still remember the name of your favorite soft toy. The primary function of a traditional (dumb) teddy bear is to provide comfort to a child. A smart connected bear could be fitted with a range of sensors: a small camera in its eye, a microphone in its ear, and a speaker in its mouth. Set aside the obvious privacy issues this raises for now. We will address that concern in a moment. Focus first on the new possibilities this creates. If the bear's camera captures the image of a book page, optical character recognition and text-to-voice software will allow the bear to read the words aloud. Our smart teddy bear can now read a book to a child. Bedtime stories are an important time for parents to bond with their children. The smart bear would not replace this vital interaction. But our bear might offer a child something that his or her parents cannot. The bear could use a cloud-based language translation algorithm to translate text into another language. A book written in English could be read aloud to your child in French. Or Mandarin Chinese. Or Arabic. Or whatever language you might choose. Klingon, perhaps? Children can now learn new languages from the bear. The bear becomes a service portal through which language training services are sold.
Back to the privacy question. The bear's camera system could be designed with the camera and character recognition software in a completely closed system. No images would ever be stored, and the sealed camera subsystem would only output text, no images. Privacy could be fully protected. Again, though, this is offered as a thought exercise, not as a product proposal.
Our bear can now teach new languages, skills that will improve the quality of a child's life. A parent is no longer buying a product. At a minimum, they are buying language-translation services through the bear. Ultimately, if their child sticks with it, what the parent is buying is a transformation, a transformation of their child into a linguist. Any economics major will explain that you can charge exponentially more as you travel up the classic hierarchy of raw materials, piece parts, products, services, experiences, and ultimately transformations. The bear illustrates the potential for IoT technology to help businesses make more money by delivering more value and capability to customers.
How much is the smart bear worth? $199? $499? More? Rather than sell the bear to you for $199, a manufacturer might let it go for the low, low price of $49. The catch: you have to sign up for the basic English-language reading service at $3.99 per month. For languages services, you can choose the European language pack that includes French, German, Spanish, and Italian for another $4.99 per month. Chinese, Japanese, and Arabic might be available at $1.99 each per month. Klingon could be thrown in for free if you buy any two language packs. You get the idea. Smart objects create the potential for new pricing models.
Our smart bear could offer a range of other services beyond reading books. It could tell a child the weather forecast, teach them simple mathematics, or answer simple questions about the world. It could track a child's speech and reading development over time. A smart, connected object can be a portal to multiple services.
Think about which new services, experiences, or transformations you could sell to your customers through service portals of your own. Think about which new customers you might reach with such devices.
A service portal initially built to solve one problem can later be provisioned with a wide range of additional services that monetize the initial investment. It's important to establish an early beachhead position for smart objects, perhaps by offering heavily discounted initial hardware. Early winners can monetize their investment by selling a suite of services and charging access fees to others. For example, if toy company Fisher Price built a popular book-reading bear, they could sell access to the bear to Johnson & Johnson. A parent might be able to order a new box of diapers by talking to the bear, and Fisher Price might take a cut of that sale. The smart bear might gain an entire app store. Apple charges 30% for all the apps that go through its app store; forward-looking manufacturers may be able to create similar money-making machines with service portals of their own.
Selling access to your customers can be a dicey proposition. It raises the critical issues of trust and privacy, a topic that's especially sensitive when you're talking about children. In part, I picked the example of the smart bear as a way to highlight the issue. Any business that develops smart objects must guard their customers’ privacy and ensure that new services are offered on an “opt-in” basis. Customers should always remain in control. If you provide enough value, are clear on the benefits you're offering, and are 100% transparent on the data you're asking customers to share in return for value, customers will generally sign up. Fall short on any of these and you're asking for trouble.
Our smart bear concept is not far-fetched. Connected toy company, Cognitoys, has already built a successful service portal. The Cognitoys Dinosaur is an educational toy designed for kids aged between five and nine years old. There's no screen and no camera. Just an eight-inch-tall plastic dinosaur called Dino with a push-button belly. Dino tells stories, has simple conversations, cracks age-appropriate jokes, teaches mathematics, plays games, and even remembers your child's favorite color. If Dino senses that a child is scared or sad, it will encourage them to talk to an adult and attempt to cheer them up with a joke. It's an impressive, but relatively simple piece of equipment. If you were to take a hammer to Dino and sift through his shattered debris, you'd find a button, a microphone, a speaker, and a fairly basic computer with integrated Wi-Fi connectivity. These components are encased in a piece of green dinosaur-shaped plastic. The hardware is simple and relatively inexpensive; the bill of materials might be no more than 10 dollars. The magic of Dino is not physical. All the clever conversational software resides on IBM's Watson cognitive computing service, built in the cloud. Cognitoys’ investment is highly scalable. Digital value scales at near-zero marginal cost. If the majority of the value you deliver is digital, marginal cost is low and high volume quickly leads to high margins as development costs are recouped. Dino's digital component could scale to many other toys: a talking fire engine, a talking unicorn, or a talking backpack.
Smart, connected objects transform business models. Let's review real-world examples from the healthcare and insurance industries.
Beam Dental created a smart, connected toothbrush that logs how often and how effectively you brush your teeth. It's like a Fitbit for your mouth. What might sound like a novelty item actually has the potential to be a major disruptor in dental insurance.
People with good dental health need fewer expensive dental interventions like fillings, root canals, and all the other painful stuff that fills our nightmares. Because Beam Dental can remotely monitor each patient's brushing habits and see that they brush their teeth properly morning and night, they know they are a lower risk. That reduced risk allows Beam Dental to offer dental insurance premiums that they claim are 10–25% lower than their competition. That's a breakthrough. Beam Dental doesn't sell the smart toothbrush as a standalone item. Importantly, it's offered as part of a full dental health service that includes as much toothpaste and floss as you need and personalized oral health tips delivered through a companion app. Beam Dental has built their business model around reducing their risk by increasing their customer's use of preventative dental care measures. This is a revolution in thinking for the dental industry. Beam Dental customers are buying an entire dental wellness system that includes the smart toothbrush, dental hygiene supplies, remote monitoring, insurance, and dental health services. How can you repackage your offerings in new ways?
Semioticons, a research consortium, is developing a home health diagnostic product named the Wize Mirror with funding from the European Union Commissions on ICT for health. This high-tech mirror bristles with sensors—including 3-D scanners, multispectral cameras, and gas sensors—to measure the general health of its users. The smart mirror monitors facial expressions to look for signs of stress, anxiety, or depression. 3-D scanners measure the shape of a person's face and track it over time, a good proxy for weight loss or gain. The mirror reviews complexions, looking for pale, jaundiced, or flushed faces. Multispectral cameras measure heart rate and hemoglobin levels in the blood. A gas analyzer measures the breath to monitor blood sugar levels, drinking, and smoking. Each health check takes about a minute. Results are displayed on the mirror, which also acts as a screen. Ultimately, the mirror will provide health and lifestyle advice based on its observations.
Insurance companies use sensors to measure rather than model risk. Typically, auto insurance premiums reflect the risk each customer presents based on their driving record, address, gender, age, occupation, and car model. From these data points, algorithms model the statistical chance of a claim, and premiums are set accordingly. Progressive's Snapshot product is a small wireless device that plugs into a standard connector in your car, usually found underneath the dashboard near the steering column. The connector shares real-time data about your steering, acceleration, and braking. Snapshot, which is about half the size of a packet of cigarettes, monitors your driving habits and measures your risk as a driver, rather than modeling it based on historical statistical data. Progressive rewards careful drivers with lower premiums and claims that premiums for customers using Snapshot are up to 30% cheaper.
American Family, another U.S.-based insurance company, formed an alliance with Alphabet's Nest division to better manage their home insurance risk. American Family provides a Nest Protect smoke detector to their home insurance customers. These smoke detectors can be monitored remotely. Customers that regularly replace back-up batteries in their smoke detector and maintain the smoke monitoring service are rewarded with lower premiums. American Family knows the detectors are in good working order, and thus the risk of undetected fire is lower.
Service portals, in the form of smart objects, will create a whole new set of business models and will birth millions of exciting new services. As the price of computing and connectivity continues to drop, expect many smart, connected objects to inhabit our lives: connected washing machines, pill bottles, pillows, shoes, rings, cars, exercise equipment, wine bottles, staplers, dog collars, collection boxes, parking meters, garbage cans, and so on.
Every business should consider their service portal strategy. Portals can deliver services to customers, employees, and business partners. Market leaders will build their own portals, while followers will strike deals to deliver services through the portals of others. What will your service portal strategy be? Just because your business hasn't built such devices in the past is no good reason for not building them in the future.
Consider a wide spectrum of possible smart objects. On one end of the spectrum are objects that are 100% physical and that have no digital footprint: “dumb” objects. At the other extreme of the spectrum are objects where the value being delivered is almost 100% digital. They have little or no physical form, for example, a QR code printed on a newspaper page that links you to an app, a web page, or some other kind of digital value. Most smart objects exist somewhere in the middle two-thirds of the spectrum, some more focused toward physical value and others more focused on digital. The Cognitoys Dinosaur is closer to the digital end of the spectrum since its physical characteristics are mostly unrelated to the value delivered through it. The dinosaur could perform its function just as well in the form of a fluffy blue cube, a mirror, or any form appealing to a child. Smart objects closer to the physical end of the spectrum have digital value that is tightly connected to their physical function and form. For example, high-end cars now include smart headlights that use sensors and digital intelligence to direct their light beams around oncoming traffic. Drivers can keep their high beams on at all times without dazzling other drivers. The digital value of these smart headlights is tightly coupled with their physical form.
My friend Dr. David Rose works at MIT's Media Lab. David's team created several objects that illustrate this idea of digital value being deeply embedded into an object's physical form and function. He calls these objects “enchanted objects.” A connected umbrella has a handle that glows when the local weather forecast calls for rain. A connected billfold wallet links to its owner's bank account. A variable-strength hinge makes the wallet harder and harder to open as the person's bank account dwindles, giving them tangible and immediate feedback on their spending ability. In both of these cases, digital value enhances the physical value provided by the object and is linked to the object's function.
The Internet of Things allows us to build intelligent infrastructure. Sensors gather data from the physical world; the output from those sensors is interpreted in the digital world, and then action is taken based on that interpretation. This process of “sense, interpret, act” occurs in a never-ending loop, known as a control loop. Decision logic makes sense of the data and determines whether any action should be taken. Actions are taken by algorithms (perhaps an AI), autonomous machines (perhaps a robot), or humans. For example:
In this way, the digital world is given limited agency over the physical world. It makes decisions based on prewritten rules created by humans.
Let's consider how IoT technology might make a garden sprinkler system more intelligent. Rather than operating on a simple timer, our sprinkler system uses a moisture sensor embedded in the soil. A small computer in the sprinkler interprets the data from the moisture sensor and compares it against a predetermined threshold value. When the threshold level is reached, signaling that the soil is dry and needs watering, the computer activates the sprinkler. A smarter system, connected to the local weather forecast via an internet connection, would only trigger if rain isn't forecast within the hour.
Control loops can be very simple, as for our garden sprinkler system, or they can be very complex. Complex control loops are used to automate manufacturing lines, control nuclear power stations, and optimize traffic flow in major cities. Some control loops, like our sprinkler, might be navigated just a couple of times a day—on, off, on, off. Other control loops occur hundreds of times a second. The navigation system on an autonomous drone measures its position in space several hundred times a second, making tiny adjustments to the speed of its rotors to maintain its course and speed under blustery wind conditions. Some industrial control applications operate with control loops that occur thousands of times a second. Industrial turbines measure internal conditions and adjust gas intake and turbine speed to optimize fuel consumption and performance.
Many IoT applications are fundamentally about the optimization of control loops. Each time, you sense the world and then make a decision. That decision is usually an attempt to optimize for some kind of result. The seemingly simple, but vitally important, question is: What result should we optimize for?
Our sprinkler system is currently being optimized around two primary factors:
To illustrate the point, let's have some fun and imagine an unnecessarily sophisticated sprinkler system. First, let's optimize our sprinkler to minimize inconvenience to people enjoying or working in the garden. Many of us remember a time in our life when we were happily minding our own business, relaxing in the sunshine on a lovely lawn, only to be surprised by a torrent of water from an ill-timed sprinkler. After the swearing is over and we've dashed to safety, we inevitably wonder if somebody, somewhere designed the sprinkler system to do that to us on purpose.
A simpler design might only water at night. A better system would use a smart camera to sense when the garden is not in use. The ultimate (and entirely over the top) system could mount cameras in the garden, or on a drone, and use image recognition technology to understand which species of plants are located where. The watering system would optimize water delivery for each individual plant and monitor plant health over time.
The point of this exercise is to illustrate that control loops can become as complicated and sophisticated as you want (and can afford) and that the key is to decide what end results you are optimizing for.
For 30 or 40 years, businesses used computers to optimize productivity and efficiency. Computers boosted office productivity and computer control made factories more efficient and boosted quality by bringing precision and repeatability to business processes. Most of the productivity gains that can be had from traditional client-server computing and automation have already been realized. Any remaining improvements will be incremental and are unlikely to be revolutionary.
The Internet of Things, coupled with cloud computing, Blockchain technology, and artificial intelligence, will lead to huge breakthroughs in business process improvement and product innovation in the coming decade. Automation will yield further advances in productivity, efficiency, and sustainability. The capabilities of the Internet of Things reach far beyond those of traditional client-server computing. The IoT creates far more intimate connections between the digital and physical worlds. The more intimately we connect the digital with the physical, the more value and capability can flow from the exponentially improving digital world into the physical world that we inhabit.
What we optimize for really matters. With IoT, system engineers can set higher goals for the automated IT systems that they build. We explore this notion more fully in Chapter 7 when we review the strategy and philosophy of automation.
Powerful computers can ingest vast amounts of complex information gathered from a wide array of disparate sources. Algorithms spot patterns in this data and make informed decisions accordingly. Often these patterns are so complex, or spread across such vast troves of data, that they are impossible for humans to see. The class of software that finds these patterns and helps us make sense of complex data is known as analytics. Analytics are key to unlocking the full value of the IoT and business process automation.
Analytics guide the operations of many businesses. Analytics are used to design election campaigns, make Netflix and Spotify recommendations, assess credit worthiness, and set the pricing of flights and hotels. In the coming decades, turbocharged by artificial intelligence, analytics will become the “brains” behind many business operations. Smart sensors will understand what's happening in the business in real time, and analytics tools will make well-informed, lightning-fast, data-driven business decisions in response.
There are four main categories of analytics. Each answers a fundamentally different question. These questions are: “What happened?,” “Why did it happen?,” “What may happen?,” and “What should we do?” The answers to these questions have tremendous business value. Let's explore each of these types of analytics in turn to understand how they can be used to make business operations run more smoothly and efficiently.
Descriptive analytics analyze historical data to help us understand the answer to the question “What happened?” They typically offer insight into the operation of complex systems: the flow of a city's transport network, the performance of a jet engine, or the operation of a large call center. Descriptive analytics turn raw data into information and insights that help people to make decisions, for example, highlighting rush-hour traffic and accident hotspots to help city planners redesign interchanges and improve public safety.
Diagnostic analytics dig deeper than descriptive analytics and help people answer “Why did it happen?” Digital marketers use diagnostic analytics to crunch through social media data—posts, shares, likes, tweets, retweets, and mentions—to figure out which messages, channels, or campaign strategies worked best on previous marketing campaigns. Logistics companies use diagnostic analytics to understand their delivery efficiency. UPS analyzed the efficiency of the routes their truck drivers were taking using diagnostic analytics. They revealed that routes that require trucks to take fewer left-hand turns are generally more efficient. As a result, UPS route-mapping software now chooses routes that minimize the number of left-hand turns needed.
Predictive analytics help you understand “What may happen?” This software attempts to predict what is likely to happen next, based on a deep understanding of what happened before. Predictive analytics have many powerful business applications. Restaurant chains use analytics to predict how busy each restaurant is likely to be on an hourly basis. These insights inform staffing levels and shift timing. Obvious data sources for these analytics include historical demand patterns and an annual events calendar. A restaurant is likely to be busier on Valentine's Day and St. Patrick's Day, but quieter during Ramadan and Lent or on a Tuesday night early in the new year. Predictive analytics factor in weather forecasts and scour social media and the web to ingest the event schedules of nearby entertainment venues, and correlate events against previous demand spikes.
Predictive analytics finds patterns of human behavior that are hidden inside massive sets of data. Walmart used predictive analytics to correlate local weather data with sales data from checkouts at every one of their stores. Walmart data scientists found that under a specific set of weather conditions, people visited the store looking to buy berries. Analytics software revealed a surge in demand whenever the weather was clear, sunny, with a light breeze and temperatures below 80°F (27°C). Acting on this insight, Walmart positions berries more prominently and boosts advertising when these weather conditions are predicted. It reduces stocks in unfavorable conditions. Incredibly, Walmart saw berry sales triple in some locations. A 10% sales uplift is a big deal in retail; a 300% boost is monumental! Walmart's analytics discovered that people are more likely to buy steak when the wind is above average, it's not raining, and the temperature is warm, but not hot. Ground beef, presumably to make burgers, sells better in warmer temperatures, low winds, and mostly sunny weather. When they spot these forecasted conditions, Walmart analytics automatically trigger the release of local ads focused on hamburgers. That results in a spike in ground beef sales of up to 18%! Understanding subtle patterns in human behavior with predictive analytics has a tremendous impact on sales for a major retailer like Walmart.
Prescriptive analytics build on all the previous levels of analytics to answer the question “What should we do?” This is perhaps the most powerful of all the four main types of analytics as it begins to shift decision-making into software. In the coming decades, prescriptive analytics, fed by vast troves of data, will quietly, efficiently make many of the decisions that underpin the world that we live in. Businesses use prescriptive analytics to help them design products, set pricing, smooth business processes, and optimize the efficiency and effectiveness of their employees by guiding them on what tasks to do next. Italy's largest bank, UniCredit, uses prescriptive analytics to make better lending decisions and optimize its capital. Oil and gas companies use prescriptive analytics to help them decide where to frack and how to optimize the fracking process. Aurora Healthcare, an organization that spans 15 hospitals, employs 6,300 people, and serves 1.2 million customers, analyzed 10 years of clinical treatment data and patient outcomes so that they could offer near-real-time treatment recommendations to their doctors based on a patient's diagnosis and history. As a result, they were able to improve healthcare outcomes and reduce readmission rates by 10%, saving $6 million each year (Source: Datafloq).
Every business should have an analytics strategy and find ways to use analytics to help them make higher-quality decisions in every business process.
When applied to people, Internet of Things technology enables us to enhance our capabilities and quantify our personal health and activity. Popular wearables focused on health and fitness measure activity and create feedback loops that encourage healthy behaviors.
Wearable technology will spread beyond the wrist and find a place on our heads, in our ears, on our fingers, and ultimately inside our bodies. Augmented and virtual reality headsets are discussed in Chapter 5, so we will skip them here.
A hearable is a smart, connected device that fits on or in your ear. Some wireless headphones already connect to voice assistant services. Future hearables will combine microphones with earphones to augment our hearing. Intelligent hearables will process sound and optimize it for us, protecting our ears in loud environments, applying personalized audio equalization (bass, treble, and mid), and filtering noise. Smart noise cancellation offers the prospect of being able to selectively remove sounds we don't wish to hear—babies crying at the mall, the rumble of an office air conditioner, or the screech of a subway train—while preserving the sounds we don't want to miss. Hearables may also enhance hearing for those who need it: first responders listening for survivors, hunters tracking animals, law enforcement chasing bad guys, or people at the back of a room listening to under-amplified public speakers. Hearables might also offer the ability to replay the last 10 seconds of sound in case you missed something, or to edit the voices of friends and family. Who wouldn't want to make their boss's voice sound like Darth Vader? Hearables will also translate languages in near real time. Waverly Labs’ Pilot in-ear wireless headphones can translate 16 languages in near real-time with impressive accuracy.
For those people who have ever found themselves shouting directly into the ear of people next to them at loud concerts, hearables offer the prospect of being connected into personal audio networks. Temporary, private audio networks could be formed that capture speech with bone-conduction microphones in the ear and transfer them to friends who can hear you no matter the volume of the death metal band on stage. Teamwork at the supermarket would also get easier, too. “Hey honey, please grab some bread and ham and meet me in the cereal aisle.”
Some hearables will measure biometric information from our ears, gathering heart rate, blood pressure, temperature, ECG, and even blood oxygen levels. Hearable accelerometers can measure movement and activity. Biometrics can be useful for authentication and security. Your heart rhythm, gait, and ear shape form a unique signature that could be used to authenticate access to devices and services, much the same way that an Apple Watch can sign you into a MacBook. NEC claims they can measure the reflection of sound waves in the ear canal and determine the unique shape of an individual ear with greater than 99% accuracy. When combined with other biometric signatures to further boost accuracy, this could provide a new root of trust for security applications that require continuous authentication. Goodbye and good riddance, passwords.
Advances in printed sensors and electronics deployed on flexible substrates will open up new platforms for wearable and medical devices in the coming decade. Future wearables will make today's products seem quaint by comparison. Researchers at MindX are hard at work developing mind-reading smart glasses that sense eye movement and brain waves to figure out where you're looking and what you're thinking. They could offer a new interface and be particularly useful for performing visual searches. Sana has built soon-to-be-FDA-certified smart goggles that promote relaxation and eliminate chronic pain using patterns of sounds and light to modulate brain waves. The goggles monitor brain activity in a closed-loop biofeedback system that learns how to reduce or eliminate pain when worn for as little as 16 minutes. Clinical trials are under way for the treatment of fibromyalgia, neuropathic pain, oncology pain, opioid misuse disorder, and severe pain.
Future wearables that combine advanced materials, sensors, and artificial intelligence to perform super-sensing will redefine the wearable landscape.
Every organization should have a comprehensive sensor strategy. The first step to building any automation strategy is to create a sensor strategy. If you don't have a sensor strategy, start today. Get eyes on your business operations and gather the data you need to automate or semi-automate decisions.
Use sensors to detect important business events so that you can act on them in near real time, speeding operations and delighting customers in the process. For centuries, bells attached to the doors of stores have alerted shopkeepers to the arrival of customers. Modern sensors can detect important business events that require immediate attention: a restaurant customer's drink being empty, a fire caused by a lightning strike, a package clearing customs, or a PR disaster unfolding on Twitter.
Starbucks in China uses the store's Wi-Fi hotspot as a sensor. Wi-Fi hotspots regularly sniff the airwaves for devices that want to connect to them. Even devices that don't connect share limited information about themselves, a unique identifier known as a MAC address. Starbucks uses the Wi-Fi hotspot to count the number of phones it can see. Most people have one phone, so the phone count is a good proxy for the number of customers in the store at any given time. Starbucks uses this information to control the music system in the store. When the store is quiet, they play more relaxing music designed to encourage customers to linger longer and maybe order a second drink or a slice of cake. When the store is busy, they increase the tempo and volume of the music, hoping to encourage people to grab and go. This approach has had a material impact on their business.
Most business events can be detected with the right sensor strategy. Detect as many business events as you can—customer-related events, operational events, employee-related events—and make your business more responsive and more efficient as a result.