Figure 8.1: What the Google self-driving car “sees” (Source: Google Self-Driving Car Project)
“I think it will be quite unusual to see cars that don’t have full autonomy... Any cars that are being made that don’t have full autonomy will have negative value. It will be like owning a horse. You will only be owning it for sentimental reasons.”
Elon Musk, CEO of Tesla, Tesla Earnings call, November 2015
At the CES in Las Vegas in 2015, Mercedes launched a self-driving car called the Mercedes F015. In fact, it reportedly drove itself to the venue. Where it differed from the likes of Google’s self-driving car, or those from Volvo, Audi or Tesla, is that it was specifically designed as a “third place” for consumers, as Dieter Zetsch, head of the German company, put it at the time. If our home is our first place and our office or school our second place, then where is the next place we spend a large portion of our time? Our car is the next logical place or space to personalise as a living space, especially once autonomous.
Zetsch described the redesign of the space in the Mercedes F015 as a return to the days of the “carriage”. In the days of the horse and cart, you would sit in the back in comfort while the driver did all the work up front. The same will be true for autonomous vehicles. In the past, driving in a car required focus and attention on driving. The vehicle’s utility was all about safety, navigation and the ability to get from point A to point B. However, as self-driving cars evolve, the requirement for the car to be driven or for the driver to focus on driving will disappear or become minimalised. As a result, the need to have a car designed around the act of driving will probably be trumped by the need to design the “carriage” space and how you utilise it when you are being driven.
Despite the suggestion of self-driving cars being considered fantastical by some, we may actually be much closer to that reality than many anticipate. Elon Musk, the CEO of Tesla, believes we are much closer to that future.
“We’re going to end up with complete autonomy, and I think [Tesla] will have complete autonomy in approximately two years.”
Elon Musk, from an interview with Fortune magazine,
21st December 2015
Google’s self-driving cars have now accumulated close to 2 million driving miles (autonomous and manual driving combined) without causing a single incident, accident or fatality.1 A Google self-driving car has been pulled over by police, although it somehow avoided getting a ticket.2 An average American driver is likely to have an accident every ten years or so, or about once in every 165,000 miles.3 So Google is already more than ten times safer than the average human driver on a purely statistical basis. Data gleaned from the Google self-driving car project has yielded other critical observations, too.
When comparing autonomous driving patterns with those of drivers in control of Google’s vehicles, Google found that when a human was behind the wheel, Google’s cars accelerated, cornered and braked more sharply than when the car was piloting itself. Other data showed that the autonomous software was much better and more consistent at maintaining a safe distance from the vehicle ahead.
“We’re spending less time in near-collision states,” said Chris Urmson, the leader of Google’s autonomous car project at a robotics conference in 2013. “Our car is driving more smoothly and more safely than our trained professional drivers.”
Google’s self-driving car has the most data publicly available about this incredible autonomous capability, but other car manufacturers like Tesla, Audi, BMW, Mercedes and Volvo all say similar things about the future of driving. Autonomous vehicles will most likely be significantly safer than those driven by humans within a decade or so.
Giving further insight into this technology, Google has disclosed that the sensors on the Google self-driving car capture nearly 1 gigabyte (GB) of sensor data every second, and subsequently process that information to identify risks or anticipate issues that it may need to react to. It reads traffic all around the vehicle, including any moving hazards on the road that could obstruct its path, it tracks the lanes and shape of the road, referring to GPS data and available map information, as well as any signage or other information that determines the speed the car can travel or where it has to respond to road conditions. It can even isolate a cigarette butt thrown on the side of the road, or a child running in front of the vehicle while chasing a ball onto the street.
The current generation of Google’s self-driving car has a Velodyne 64-beam laser embedded in the nose of the vehicle, which generates a 3D map of everything it sees around it. The on-board computers combine this with high-resolution maps of the world, historical driving data and camera and radar images. The vehicle carries four radars (two at the front, two at the rear), a camera that detects traffic lights, a GPS, a wheel encoder and an inertial guidance and measurement unit.
The computers powering these autonomous vehicles are not particularly large either. About the size of an old laptop, they typically fit in the boot of a regular vehicle. In 2014, Audi released a computer called zentrales Fahrer Assistenz Steuergerät (zFAS), or central driver assistance control unit, that is small enough to be mounted in the rear quarter of the vehicle, accessible from within the boot. The zFAS computer powers the Audi autonomous race vehicles Ajay and Bobby that we discussed in chapter 3.
The reaction of many diehard drivers, of course, is to say that they’ll “never trust a self-driving car to drive them”. Given the facts though, that argument is just as illogical as the one used by people who refuse to fly in aircraft and prefer to drive. Statistically, flying is significantly safer than driving a car. It has often been said that you have a greater risk of being killed on the way to the airport than in flight. Put another way, the odds of dying in a motor vehicle accident are 1 in 98 spread over your lifetime, but 1 in 7,178 for commercial air transport.4 Those are pretty good odds, odds that show that autonomous piloting is statistically very safe.
For 90 per cent of the time that you spend in a modern commercial aircraft, an autopilot computer is flying it.5 Many of the newer aircraft flying today have auto landing systems (ALS) that even allow them to land at designated airports unaided by a pilot. While ALS systems are rarely used, redundant systems, better training and simulators, autopilots and modern navigation aids have all made flying the safest form of mass transportation known to man. So it may be argued that the same autopilot capabilities when applied to driving might have similar results.
We’ve managed to improve car safety systems considerably with seatbelts, rigid passenger shells, airbag deployment, safety glass, anti-lock braking systems (ABSs), collision avoidance systems and other such capabilities. Yet in the United States, Australia, Canada, Germany and the United Kingdom, there are still an average of around five to ten deaths for every 100,000 vehicles, or about one death for every 100 million miles driven. We can combat drink driving and put more controls on speeds and road conditions, but the reality is that we can’t significantly reduce the number of vehicle deaths further because of the remaining human factor. While fatalities have been declining, they’ve been doing so at a slower and slower rate. Indeed, the biggest risk we have when it comes to driving these days is simply human error. To emphasise the fragility of these statistical gains previously made, it is estimated that nearly 25 per cent of all accidents in the United States in the last four years were caused by people texting while driving. As a result of such factors, since 2011, fatalities have started to increase slightly again, bucking decade-long declines.
Technology is distracting us, and the only way to combat that is to either raise awareness and ban the use of phones while driving (which has had limited success) or introduce autonomous vehicles so that our use of technology is no longer a factor.
Taking all of this into account, it turns out that machine intelligence is going to be able to demonstrate fairly quickly that it is better and safer at driving than us humans. In fact, based on Google’s beta testing of its self-driving cars, the existing units are about ten times safer than human drivers. For every 1 million miles that the existing fleet of self-driving cars undertakes, without causing an accident, we will see that number effectively double. Probability means that at some point a self-driving car will cause an accident, and that at some point it will probably be involved in a fatality, but these self-driving cars will still be demonstrably safer than human drivers.
As Brad Templeton, a Singularity University professor who worked with Google on the self-driving car, articulated to me during a recent interview:
“Self-driving cars don’t get tired, don’t get drunk, don’t get distracted, don’t get road rage and don’t need a rest, unless it might be to charge.”
Brad Templeton, Singularity University,
author interview in May 2015
Within ten, or even five, years in developed economies, semi-autonomous or self-driving cars will probably be commonplace. Think about that for a moment. In about the same time that it has taken for the iPhone and smartphones to dominate every corner of society, we will see smart, autonomous cars exploding onto the scene. Business Insider has estimated that we’ll have 10 million cars with self-driving features on the road by 2020. The exponential curve of this technology means that there will be close to 100 million self-driving cars on the road just ten years after that.
Within 15 years, we can expect that major cities and local authorities will be giving strong preferences to self-driving cars. Within 20 years, cities like London and New York won’t just have congestion charges, there will also be charges for traditional, human piloted vehicles to enter the city centres, or more probably even banning them from city streets. While there will be some protests against banning “human drivers”, remember that the generations making these governance decisions will be Gen Y and Gen Z, not the likes of the baby boomers who grew up with gas guzzlers, V8 engines and the oil boom. Our kids expect this to happen and will be quite happy to trade in the art of driving for more screen time in an autonomous vehicle.
Elon Musk said on a call before the 2016 Detroit Motor Show that within 24 to 36 months, you’ll even be able to summon a car to drive itself across the country to meet you:
“If you’re in New York and the car’s in Los Angeles, you can summon your car to you from your phone and tell the car to find you, and it’ll automatically charge itself along the journey.”
Although the current barrier of entry to autonomous vehicles is cost, that cost is rapidly shrinking. In 2010, the cost of Google’s self-driving technology was around US$150,000, of which US$70,000 was just for the highly accurate laser-based radar, or light detection and radar sensor (lidar). German supplier Ibeo, which manufactures vehicular lidar systems, claims that it will be able to mass-produce them by 2017 for as little as US$250 per vehicle. Computational processing is probably another large component of the overall price, and it has a long history of exponential cost reduction. In addition, we can expect battery technology to improve dramatically and allow self-driving vehicles to eventually get through an entire day without the need for a charge.
The sharing economy for vehicles will thrive. Electric, self-driving vehicles will be significantly cheaper to run, but the average city dweller won’t even need to own a car, or they may choose to own a share in a vehicle. The user will simply rent a vehicle by the hour, and each vehicle will shuffle between owners or renters themselves, stopping to charge at the required charging station in quiet periods.
The trend of young adults moving away from vehicle ownership is already becoming evident. Instead of asking for their own car when they reach driving age, teens are now asking their parents for an Uber account.6 So this is not just a factor of electric, self-driving cars; the sharing economy is already starting to shift behaviour towards dramatically different vehicle ownership models. Children who have grown up with parents who use Uber or ride-sharing services will do the maths and find that it is cheaper to not own their own vehicle in an average city with good public transportation and an autonomous vehicle network.
For those with a commute, this is where the Mercedes vision of the self-driving car gives us a glimpse of the near future. Realising that a self-driving car does not need to be optimised for driving, the interior space could instead be used for entertainment, eating your breakfast on the way to the office, as an office itself or just as an extension of your personal space. How will users of self-driving cars personalise their vehicles? The future customisation of a self-driving car may be along the lines of how you personalise your home today, instead of how you may have personalised a car in the past. The choice of entertainment system, seating, display tech, even a 3D-printer food processing unit could be all the rage in this new personal space. Maybe living in your self-driving car might be fashionable for entrepreneurs, making it a sort of mobile home/office.
The only thing that might hold all of this back, conceivably, is legislation. However, once manufacturers demonstrate the safety of self-driving car systems, it is far more likely that both passengers and legislators will start to opt in voluntarily. Some legislators will insist that autonomous vehicles must have the option for a human to “take over”, and there will no doubt be purists who try to hack around autonomous routines in some way. Moreover, we can expect lobby groups of manufacturers who fall behind in autonomous technology to attempt to muddy the waters with figures around safety. The first death of a passenger or a pedestrian by an autonomous vehicle will be a watershed moment. It is unlikely, however, to stop autonomous cars from dominating our future. Interestingly, the CEO of Volvo, Håkan Samuelsson, has already said that Volvo will accept liability when a self-driving car is involved in an accident.7 That is a big deal!
Early applications of self-driving technology will likely be commercial. Self-driving delivery trucks combined with either drones or delivery robots (to drop a package at your door) will be much cheaper to run than the human-operated FedEx and postal service trucks that we see on the roads today. The delivery of shipping containers to ports, and other such transportation, will also quickly become autonomous.
While Uber is investing heavily in this area, there may still be some humans in the driving services business for some time. Initially, the wealthy will express their wealth by owning a self-driving car, but there will most likely be a U-turn at some point in the future, where having a human driver becomes an expression of wealth, and not the other way around. For transportation of goods, however, no such humanity is required or beneficial in the medium term. Self-driving cars will lower cost, injuries and fatalities as well as increase the utility of transportation networks considerably. Particularly if those autonomous vehicles are electric vehicles.
For those of us who regularly commute today, with self-driving cars we’ll suddenly have a great deal more time on our hands, and that will change the way we think of driving in fundamental ways. The family road trip might take on a whole new meaning. So the logical next question is: how long will it be before we have flying cars like the one Doc drove/flew in Back to the Future?
Flying car enthusiasts may recognise the name Paul Moller. In 1974, Paul Moller touted a flying saucer type car he called the Discojet, raising funds for the project. In 2003, Moller was fined by the Securities Exchange Commission (SEC) for making false and misleading statements to investors. As the SEC put it: “As of late 2002, Moller International’s approximately 40 years of development has resulted in a prototype Skycar capable of hovering about fifteen feet above the ground.” In 2013, Moller launched a crowdfunding campaign on Indiegogo8 to raise a further US$958,000 for the project. It closed in January 2014, having raised just US$29,429.
We have been dreaming of flying cars since the 1950s.
When the U.S. civil aeronautics administration certified the Aerocar for operation in 1956, it seemed inevitable, at least to aerospace engineers, that before long the flying car would take its place as a fixture in the garage of the typical suburban ranch home. Yet that was not to be. The Aerocar, which looked like a car but had wings and could take off on a short runway, was too expensive to justify mass production. Aerocar International built only six of these vehicles, leaving the promise of the flying car unfulfilled—except in episodes of The Jetsons.
“The Flying Car Will Finally Fly—and Drive,”
Scientific American, January 2013
Today, however, at least two companies have made flying cars a reality. The Terrafugia Transition and the Aeromobil flying car have both successfully made the transition from car to flight and carry actual passengers.
The Aeromobil is the more futuristic looking of the two vehicles. This flying car is capable of 125 mph (200 km/h) powered flight and has a range of 435 miles (700 km). Built in Slovakia, an unlikely destination for the future of flying cars, the Aeromobil is the result of 25 years of development. With the advance of composites and engine design, the promise of this technology has finally caught up with reality.
But don’t expect to walk down to your local Aeromobil dealership and start flying anytime soon. At best, the Aeromobil will be classified as a sports aircraft, requiring a sport pilot licence or a minimum of at least 20 hours of flight training. The Aeromobil can indeed fly, but it still needs a runway like a conventional aircraft and is subject to all the same rules as a Cessna, Cirrus or Piper aircraft. Want to fly a car? You’ll still need pilot’s wings.
That may change in the future, however. Increased computing capability has a role to play here but airspace governance is the key. Currently, separation9 of aircraft flying in either controlled or uncontrolled airspace is a function of pilot awareness, collision avoidance systems and radar vectoring by air traffic control (ATC). The Next Generation Air Transportation System (NextGen), which is emerging in the United States, could possibly set the basis for autonomous flying cars.
Between 2012 and 2025, the implementation of NextGen is expected to cost US taxpayers between US$20–25 billion but will result in travel delays in the United States being cut by more than 30 per cent, with savings well in excess of the investment. At the heart of NextGen is a technology called automatic dependent surveillance–broadcast (ADS–B). ADS–B enables an aircraft to determine its position via GPS satellite, and to broadcast that to ATC and other aircraft that can maintain separation. ATC systems currently use radar and transponders to identify aircraft. While sophisticated, these are still prone to both technical and human errors. ADS–B works like an asynchronous communications network in which aircraft are nodes, and allows aircraft movements to be coordinated in real time much more accurately than current systems. ADS–B allows aircraft to maintain separation independently in the absence of radar coverage or access to the ATC radio network. It is entirely conceivable that ADS–B could evolve into part of an autonomous network with aircraft that can fly themselves in the not too distant future.
The ADS–B network would conceivably allow the deployment of vehicles like the EHang 184, an autonomous helicopter drone that has seating for a single passenger. The drone was launched at CES in Las Vegas in 2016, and represents the first possible deployment of an autonomous flying vehicle.
Unlike self-driving cars, which are focused primarily on the comfort of passengers, it is more likely that the first self-flying vehicles won’t be passenger aircraft at all, but drones. On 1st December 2013, Amazon CEO Jeff Bezos appeared on 60 Minutes to launch an audacious plan to ship purchases to Amazon Prime customers via drone. If this had been aired on 1st April, most of the US population would have assumed it was an April fool’s joke, but Bezos was serious. He pointed out that 86 per cent of Amazon’s orders were less than 5 pounds in total weight and that orders in this category could be fulfilled in 30 minutes or less using a drone-based delivery system. Sounds like science fiction? Bezos doesn’t think so.
In a letter dated 9th June 2014 to the Federal Aviation Association (FAA), Bezos made a number of very interesting points in respect to Amazon’s development of its Amazon Prime drone fleet. He pointed out that Amazon is already deploying its ninth generation of aerial vehicle, has ex-NASA engineers (including an astronaut) on staff working on the project and that “one day, seeing Amazon Prime Air will be as normal as seeing mail trucks on the road”.
The use of drones has been toyed with in battlefields since World War II. The United States started working in earnest on drones as early as the Vietnam War, but it wasn’t until the 1982 Israeli conflict with Syria when unmanned aerial vehicles (UAVs) were used with significant success. Since then, the drones being deployed in the theatre of war have become incredibly sophisticated. On 16th April 2015, the US Navy demonstrated the ability of its X-47B unmanned carrier air vehicle demonstrator (UCAS-D) to conduct mid-air refuelling with a KC-707 tanker. The same X-47B has already demonstrated the ability to consistently land on US carriers at sea. However, this ability has not come without controversy.
Since 2004, the US government has conducted hundreds of attacks on targets in Northwestern Pakistan using UAV drones. The debate regarding civilian versus military casualties in this so-called “drone war” is significant, with recent estimates ranging from 286 to 890 civilian casualties (168 to 197 of those being children).10 The Peshawar High Court has ruled that these ongoing attacks are illegal, inhumane and violate the UN charter on human rights and constitute a “war crime”.11
The use of drone aircraft is set to explode, but in this respect while armed drones will probably remain the purview of the armed forces and perhaps police, aerial surveillance capability is now widely available to the public at large.
This afternoon, a stranger set an aerial drone into flight over my yard and beside my house near Miller Playfield. I initially mistook its noisy buzzing for a weed-whacker on this warm spring day. After several minutes, I looked out my third-story window to see a drone hovering a few feet away. My husband went to talk to the man on the sidewalk outside our home who was operating the drone with a remote control, to ask him to not fly his drone near our home. The man insisted that it is legal for him to fly an aerial drone over our yard and adjacent to our windows. He noted that the drone has a camera, which transmits images he viewed through a set of glasses.
Capitol Hill Seattle Blog, 8th May 2013
It may be that the novelty for drones wears off in the future, relegated to history like the Segway. However, given the application of drones for professional photography, recreational use and so on, it is unlikely. The FAA certainly doesn’t believe that this problem is going to go away. Consequently, it has been putting increasing effort into regulating the use of unmanned aircraft systems (UAS). In February 2015, it listed the following restrictions for people flying UAS for personal use, essentially classifying these UAS as model aircraft:
• Fly below 400 feet and remain clear of surrounding obstacles.
• Keep the aircraft within visual line of sight at all times.
• Remain well clear of and do not interfere with manned aircraft operations.
• Don’t fly within 5 miles of an airport unless you contact the airport and control tower before flying.
• Don’t fly near people or stadiums.
• Don’t fly an aircraft that weighs more than 55 lbs.
• Don’t be careless or reckless with your unmanned aircraft—you could be fined for endangering people or other aircraft.
Specifically in respect to photography or video, these FAA rules suggest that taking photos for personal use is recreational. However, for the same reason why you can’t climb a tree and photograph what your neighbours are doing in their backyard, using a drone to photograph someone on their own property is illegal, and will at some point in the near future result in a significant lawsuit being brought against a pilot of a “hobby” drone.
On 24th December 2015, the FAA announced further controls over privately owned drones, requiring all aircraft weighing more than 0.55 pounds (250 grammes) and less than 55 pounds (approximately 25 kg), including payloads such as on-board cameras, to be registered.
Criminals are using drones consistently too. In prisons across the United States, drones are now regularly being used by operators to drop contraband into prison yards. The US authorities have detected half a dozen similar attempts at corrections facilities in the past two years. In the same period, there were also reported attempts in Ireland, Britain, Australia and Canada.
Drone drops are the high-tech equivalent of smuggling a file into a prison in a birthday cake, and it underscores the headache that drones are now creating for law enforcement, who have very few ways of stopping them for now. Smartphones, drugs and smartphone chargers are the hot property for drone drops. For instance, the warden of the Lee Correctional Institute, Cecilia Reynolds, said that her officers had found 17 phones in one inmate’s cell. These prison officers suspected that the phones were delivered via drone. Prison phone calls and emails are monitored today so smartphones circumvent that monitoring. Will prisons be forced to install anti-drone-aircraft defence systems in the near future? Maybe a net over the recreation area might have to do…
As mentioned earlier in chapter 3, Facebook is developing a network of solar-powered drones codenamed Aquila (the Latin word for “eagle”) that will stay aloft for months and continuously beam Internet access via laser to millions of users on the ground. This is designed to give coverage to users in remote communities, such as in Africa where mobile coverage is scant or bandwidth non-existent. Facebook’s Aquila drone has the wingspan of a 767 aircraft but weighs less than a car. The company commenced test flights in the summer of 2015.
The use of new composites, solar power and even the resurgence of zeppelin technology could power a real renaissance in the use of the skies. Of course, it could get pretty crowded up there as a result, which is why systems like ADS–B are critical, but is also why AI in aircraft is a given in the future, if just for collision avoidance alone.
Within 50 years, the case for self-flying cars will be much stronger than it is today. The real question that remains is not whether we could have self-flying cars, because the technology for automation seems mainly achievable, but what will fuel these self-flying cars?
On 21st April 2015, a new maglev train near Mount Fuji in Japan clocked speeds of 375 mph (603 km/h). Maglev is short for “magnetic levitation” and is a technology that allows a train (or object) to be suspended above a rail with no other support than the use of magnetic fields. In this instance, the Japanese maglev train is suspended about 10 cm from the electrically charged magnets that are effectively the “track” or rail. Such a design produces a much quieter, smoother and faster ride than conventional high-speed rail.
Three years earlier in July 2012, at a PandoDaily event in Santa Monica, California, Elon Musk mentioned to an assembled group that he was thinking about a “fifth mode of transport”, calling it the Hyperloop. On 12th August 2013, Tesla and SpaceX (both companies founded by Musk) released preliminary designs for a Hyperloop Transportation system on their blog sites. Musk called this open-source design, asking others with interest to contribute to the design.
The initial proposed route for a US$6 billion passenger version of the Hyperloop ran from the Los Angeles region to the San Francisco Bay Area with an expected transit time of 35 minutes. The Hyperloop would thus traverse the 354-mile (570-km) route at an average speed of around 598 mph (962 km/h), with a top speed of 760 mph (1,220 km/h). In January 2015, Musk announced that he was building a privately funded Hyperloop test track in Texas, about 5 miles (8 km) in length, for university and private teams to test and refine transport “pod” designs. Two additional start-ups, Hyperloop Technologies and Hyperloop Transportation Technologies, are both building their own two-mile and five-mile test tracks, respectively.
The Hyperloop is a form of vactrain, or near-vacuum train. The biggest challenge traditionally in creating high-speed rail that is able to compete with air transportation systems has been the ability to circumvent friction and air resistance as speeds climb. Magnetic levitation transportation systems such as the Japanese train mentioned above use a number of very large magnets to propel the train. In fact, the JR-Maglev trains have superconducting magnetic coils. The amount of horsepower required to push a car or a train significantly above the 311 mph (500 km/h) barrier starts to climb steeply and means that if you want to get to aircraft-type speeds, say 500 mph (800 km/h), the mathematics don’t work. There’s just too much friction and air resistance.
As early as the 1960s, a proposal was put forward to build a transatlantic tunnel between New York City and London using a 3,100-mile (5,000-km) long near-vacuum tube with vactrains, or maglev trains operating in near vacuum. The system, which resembled earlier patents issued to Robert Goddard (the father of modern rocketry), was theoretically capable of allowing speeds of up to 5,000 miles (8,000 km) per hour. Meaning the transit time from New York to London would be less than an hour.
The “Musk” Hyperloop resembles these vactrain proposals but would operate at approximately one millibar of pressure, qualifying as a “partially evacuated tunnel”. Because of the low-pressure, warm air proposed for the Hyperloop steel tubes, the pods projected to travel at around 760 mph (1,220 km/h) through these steel tubes would not actually break the sound barrier. Therefore, the pods would not need to be designed for passing through the transonic phase or coping with the shock of a sonic boom.
In any respect, Elon Musk thinks that he can get you from New York to Los Angeles in around 45 minutes with this technology, so I for one would be keen to try it out!
When Nest introduced its smart thermostat, many people probably thought “so what”. However, the company quickly went on to be acquired by Google for US$3.2 billion following huge early success. In a 2014 article in Forbes, it was revealed that Nest’s digital, smart thermostat was in 1 per cent of US homes, or about 1.3 million households, that the company was selling more than 100,000 new units every month and that sales were accelerating.
It’s simple how the product works. Nest connects with the existing heating, ventilating and air-conditioning (HVAC) system in a home or office in the form of a smart thermostat that more or less anyone who can use a screwdriver can install in a few minutes. The thermostat then interfaces with the web and optimises your home heating and cooling system based on whether you are at home, what the temperature outside your home is and on peer group analysis. The average Nest user reported savings of 10 to 12 per cent on their heating bills and 15 per cent on their cooling bills,12 or an estimated average savings of between US$131 and US$145 a year. Smart thermostats are just the start, however.
In Iron Man, Tony Stark’s lab and home are interfaced with an AI called J.A.R.V.I.S. (Just A Rather Very Intelligent System) that controls all of the smart elements of Stark’s lab including security systems, power, telecommunications and even the manufacturing of Stark’s latest suits. As we’ll see in our chapter on Smart Cities, there is a definite move towards smart infrastructure as part of the Augmented Age, but smart buildings and smart homes are all part of the mix.
We talked a bit about Jibo and Amazon Echo in chapter 3, but out of the two, Amazon Echo is about the closest we have to a Star Trek-style computer (that we can talk to) in our home today (if you have one, of course).
Amazon Echo’s capability is fairly typical of the core smart house capabilities that we’ll see emerge over the next decade or two. As devices like our TVs, lights, thermostats, garages, cooking appliances, coffee machines, robot vacuum cleaners and so on will all connect to the Internet (the Internet of Things), we’ll need some basic household management capability that looks after them. These devices will be smart enough to talk to each other, but we’ll have both a management layer and an interaction layer in the home (or in the cloud).
Apple has also been working on a core capability for household management called HomeKit. Through its system, the management of your home will become largely automated. We’re not talking about just home automation technology here, as we’ve traditionally labelled this, but more an integrated home ecosystem that helps you set up and manage the smarter elements of your home.
As technology like HomeKit, Echo and individual devices become more intelligent, such resource management will have some primary goals and objectives:
1. manage the home efficiently
2. personalise your home environment around your tastes, likes, etc.
3. respond to real-time events and requests as required
4. learn
5. inform
Two key areas of the smart home that will undergo augmentation are the kitchen and the bathroom.
Despite years of talking about smart fridges that order your groceries via the Internet, the closest we’ve got is Amazon Dash Button. However, by 2030 with robotics, drone delivery and the like, you will have the ability to have your groceries and shopping delivered automatically from a smart kitchen initiated order. It’s also likely that we’ll continue to automate cooking in the kitchen. While a Star Trek-style replicator is decades away, a 3D printer that prints a burger or a pizza will be viable by 2030, if not earlier. Natural Machines launched a kickstarter campaign in 2014 for its 3D “Foodini” printer, which will be capable of printing various foods like pasta, cookies, crackers, bread, snacks, etc. A robotic chef (shown below) like the one Morley Robotics is working on is also a distinct possibility within the next 10 to 15 years.
The smart bathroom is sure to incorporate not only smarts into the bathroom mirror, but smarts into other appliances. It is quite possible that your smart toilet will be analysing your waste and looking for imbalances or evidence of emerging health conditions.
Ultimately, this technology around your home will just integrate with your personal AI. It’s one of the reasons why personalised AI assistants, what I refer to as Life Stream in chapter 7, are going to be such huge business. Google, Facebook, Apple and Amazon are all investing in this technology, but the big business will be integrating this across our homes, offices, cars and smartphones.
____________
1 There have been multiple incidents of other cars running into the back of Google’s self-driving cars at intersections, and other incidents when human drivers have been driving the cars. However, the cars have yet to have an accident while in autonomous mode.
2 The car was reportedly pulled over for driving too slowly. See Marco della Cavva, “Google self-driving car pulled over, avoids fine,” USA Today, 13 November 2015.
3 US Federal Highway Administration, AllState Insurance
4 2009 figures released by the National Safety Council
5 John Cox, “Ask the Captain: How often is autopilot engaged?” USA Today, 11 August 2014.
6 Including my 15-year-old daughter. In fact, she said instead of buying her a car I should get her an Uber account.
7 Volvo Press Release, https://www.media.volvocars.com/global/en-gb/media/pressreleases/167975/us-urged-to-establish-nationwide-federal-guidelines-for-autonomous-driving
8 https://www.indiegogo.com/projects/actually-fly-the-m400x-skycar-into-history
9 Separation refers to the distance maintained between aircraft while flying. The opposite of separation is, of course, an in-air collision, which is generally not positive.
10 Bureau of Investigative Journalism Report. October 2014 Update: US Covert Actions in Pakistan, Yemen and Somalia.
11 A. Buncombe, “Pakistani court declares US drone strikes in the country’s tribal belt illegal,” Independent, 9 May 2013.
