When journalists ask me what will be the "next big thing" to come out of Silicon Valley, i reply that it will be the "next small thing". The revolution has quietly taken place in the world of technology is the sensor revolution. Sensors have become cheap, light, small and powerful. So we can now put sensors into every object. The difference between being a rock and a living being is the sensors: living beings "sense" the environment and react to it. When we put sensors into objects, we inject a sort of life into them. They can now act on their own. And the "cloud" allows them to communicate. The cloud organizes them into an ecosystem.

There is a website that has the real-time number of sensors in the oceans: http://m1.paperblog.com/i/203/2031190/open-data-big-data-oceanografia-marinexploreo-L-SNRI3L.jpeg In 2015 there were already 750,000 high-tech sensors in the oceans of the world, of which about 45,000 are feeding data into computers, producing 1.5 billion data every second. The sensors of the 1980s were expensive and big, and probably not water-proof, and the computers of the 1980s were not powerful enough to handle so many data. Now we can put sensors everywhere, even in the ocean.

It is not trivial to write applications for the Internet of Things. We need software for monitoring sensors and processing their measurements in real time, and to broadcast the data over the cloud. Ideally, the smart objects in your house should form an ecosystem: a system of coexisting and collaborating devices. An infrastructure providing a public cloud service for machine-to-machine communication has come to maturity via "machine cloud" pioneers such as Jasper Technologies, founded in 2004 in Santa Clara by Jahangir Mohammed (and acquired by Cisco in 2016); London-based Pachube (now Xively); Boston-based Axeda (now PTC); and Minneapolis-based Spark (now Particle). Spark, for example, provides a development kit for makers and hobbyists to build smart devices and its SparkCloud to make them communicate over the Internet.

San Francisco-based SeeControl (acquired by Autodesk in 2015) provides a cloud-based platform for makers to create M2M (machine to machine) applications by connecting and managing smart sensors and devices spread around the world. Other products for connecting (wireless) sensors to the cloud come from Libelium (Spain, 2006), one of the world's pioneers; Seebo (Israel, 2012); Konekt (Chicago, 2013), that is betting on GSM instead of WiFi or Bluetooth; and Temboo (New York, 2013), that partnered with Arduino to manufacture a chip with built-in WiFi. Notice that Seebo's main customers seem to be toy manufacturers: maybe we will see children playing with interconnected toys before we see adults enjoying interconnected appliances.

Over the last few years all the big names in "cloud services" have quickly changed their brochures to advertise that their cloud can be used for the Internet of Things. If you search the web for "Internet of Things and cloud", the first names that you find are Amazon, SalesForce, IBM, Oracle, Google, Microsoft... It is not difficult to imagine that Amazon's Kinesis, Google's Cloud Dataflow, Microsoft's Azure and so forth can be easily expanded to handle thousands of devices.

Luckily the big companies have an interest in creating a community of manufacturers and programmers, so some of the software is arriving for free. In 2016 IBM transformed its big-data product Streams into an open-source software called Quarks for the manufacturers of smart objects and for the developers of applications.

The Internet of Things will produce a lot of data: who is going to do "data analytics"? All these devices will need firmware updates just like any smartphone, tablet, laptop and desktop computer: who is going to manage and direct these firmware updates? These are the functions that have been traditionally been performed by backend machines. They are now performed in the cloud. It makes sense that the same services will be used for the Internet of Things. So the software component will be as important as the hardware component.

Let's not forget who is the first beneficiary of privacy: the criminal. Privacy is much more valuable for a criminal than for me. I don't care if you search through my garage or even my bathroom: you will only find dirty clothes. But a criminal does care. Privacy looks like a value, but sometimes it is a liability because it tends to reward those who have something to hide. Many crimes have been exposed precisely because today there is less privacy, and the criminals have been caught "on tape" (by videocameras or by the smartphone of an eyewitness). My friend Antonio was able to track down the Uber driver who hit his motorcycle because Uber's computers immediately identified who was driving that car in his street at that time.

Today people talk about living longer and longer lives, and even about immortality. This is impossible to achieve in the near future. But it is possible to improve our near future in many ways, and eliminating crime should rank very high in our priorities. Villages used to have very low crime rates. Big cities created the high crime rates that we are used to. The Internet of Things can return us to the crime-free life of the old-fashioned village. Today we escort our children to school until they are almost 20 years old because we have so much crime. In the future your children will be able to go to school by themselves because all the objects between your house and the school will be programmed to make sure that nothing happens to them.

The Economist magazine in February 2016 ran an article titled "Mean Beaches" about the fact that 99% of the crimes committed in Mexico go unpunished. This encourages people to commit more crimes, not fewer.

Lithium-ion batteries were introduced in 1991, one of Japan's great inventions. Since then, their performance has improved between 5 and 10% per year. Unfortunately, our needs have improved a lot faster. Furthermore, they have not shrunk much in size. We want lighter cameras, lighter smartphones, lighter laptops, etc but these devices consume more and more electricity, no they need bigger and bigger battery packs. For 20 years mobile phones have been getting smaller and smaller. Now, suddenly, smartphones are getting bigger. Open the back of the smartphone and you'll realize why: the battery is still the same lithium-ion battery, but it is twice the size of a few years ago. The result has been a spike in the price of lithium, one of the few commodities whose price has increased over the last few years. This will also create geopolitical complications because just four chemical companies (Albermarle, with headquarters in Louisiana, Chile's SQM, FMC of Philadelphia, and China's Tianqi) account for most of the world's lithium. In 2012 the US government decided that batteries are such a strategic asset that it launched a program called "Joint Center for Energy Storage Research" (JCESR). California has a goal of 1.3 gigawatts of electricity from batteries by 2020, which is more than double what the entire world has today (0,5 gigawatts). Big companies have opened their own research centers on batteries. The Chevrolet promised that the 2017 Bolt will have batteries never seen before. Tesla and Panasonic are building a factory in Nevada to manufacture advanced car batteries. Tesla will soon start selling the Powerwall, a home battery that charges using electricity generated from solar panels. In the short term, startups like 24M (Boston) are trying to invent better and cheaper lithium-ion batteries. My main hope is in nanotech. For example, Israel-based StoreDot, founded by former SanDisk executive Doron Myersdorf, is working on "super-capacitors" that would match Tesla-Panasonic batteries but recharge in minutes instead of hours. The Internet of Things needs batteries that are light, cheap, that charge in seconds, and that last months.

Because of the limitation of batteries, the current IoT applications are mostly related to electrical appliances (objects that are physically plugged into a power outlet). In 2015 Amazon launched its Dash Replenishment Service which consists in linking "smart" appliances with its order system so that the appliances can automatically re-order supplies when they are running low. For example, in 2016 Amazon started selling a printer that re-orders ink when it is low and a washing machine that re-orders detergent when it is low,

Security on the Internet of Things is obviously not trivial. Some companies are adopting a strategy similar to what doctors do to prevent infections: check for signs of intrusion and immediately isolate compromised devices before they can infect all the other devices in the network. For example, Atlanta-based Luma sells a WiFi router capable of checking traffic inside the home network and detecting "infections", and Finnish-based F-Secure Sense sold a security monitor, a device sitting between the home router and connected devices to scan all traffic for "infections". ForeScout, originally founded in Israel in 2000, helps organizations catalog the devices that are connected to their networks, some of which could be "infiltrators" trying to camouflage as legitimate devices of the organization.

Afero, founded in 2014 in Los Altos by Danger's co-Founder Joe Britt, provides security that works even when there is no WiFi or WiFi is down, complementing it with LTE and other communication technologies.

There is also a very simple reason: the companies that will make objects and appliances "smart" will (mostly) be the ones that already make those objects and appliances; but they have an inertia, i.e. it takes time for them to change those objects.

So right now we are still connecting "things" via intermediaries. My favorite example is Waze, that combines GPS-based mapping technology with crowd-sourcing to provide live advice to drivers on which route to take. Each car using Waze sends data to Waze about traffic, and then Waze sends data to all cars about traffic. Waze acts as the intermediary between cars. The cars that use Waze are connected through Waze. In an ideal world the GPS technology would be inside the car (a lot of new cars already have built-in GPS technology) and the car would be "smart" enough to use the data from Waze directly. Our children will laugh when we tell them that in 2016 we needed a smartphone to tell us which way to go. In the future the car will know which way to go. Today the connected car platforms (Automatic in San Francisco, Dash in New York, Zubie in South Carolina, etc) are still relatively unsuccessful.

The idea of the Internet of Things is old (Mark Weiser's article in the Scientific American magazine titled "The Computer of the 21st Century" that talks about "ubiquitous computing" came out in 1991; Paul Saffo's article "Sensors - The Next Wave of Infotech Innovation" came out in 1997) but the implementation had to wait for a number of technologies to get into place: an operating system, a hardware platform and more Internet addresses. In 2003 Adam Dunkels at the Swedish Institute of Computer Science wrote an open-source operating system for the Internet of Things named Contiki. Hardware platforms such as Arduino, invented in 2005 by an international team based at Ivrea's Interaction Design Institute in Italy made smart objects practical. Arduino was meant as an open-source platform for a community of makers, facilitating the transition to a world of interactive objects that could sense and control the physical world thanks to microcontroller-based kits. The Internet Protocol (IP) is the network protocol that identifies things on the Internet. By the year 2000 the old IPv4 (published in 1981) was becoming too limited to handle the number of things connected to the Internet. Finally, in 2006 IPv6 published in 2006, introducing a 128-bit address that allows 2 to the 128th power (or 10 to the 38th power) Internet addresses, enough to assign an Internet address to every atom of the surface of the Earth. (Solve one problem and you create another one: IPv6 sounds great, but in practice its addresses are so big that they become a problem to be transmitted by small things with tiny batteries).

Chip manufacturers are rushing to design boards containing all the sensors and chips needed to wirelessly connect objects. In 2014 Samsung launched the open platform ARTIK and acquired Palo Alto-based SmartThings and then it acquired French startup Sigfox whose goal is to build a low-cost alternative to the cellular network for connecting low-battery demand objects. In 2015 Intel introduced the Curie module, comprising a sensor hub, a microprocessor and a Bluetooth connection in a tiny space.

It is interesting that in 2015 Intel made a deal with Arduino to incorporate Curie in a new cheap Arduino board, the 101. This marked another victory by the open-source world over the corporate world. The Curie may be a great chip, but Arduino is the established standard: when Intel decided to enter the IoT market, it had to partner with Arduino.

First of all, it is interesting that we are discussing privacy when discussing (smart) objects. Ten years ago we would have discussed privacy in the context of email and browsing. Five years ago the context was Google searches and Facebook posts. In the near future you will have to defend your privacy from objects.

Secondly, we already lost our privacy. It is not true that Google and Facebook stole our privacy: we gave it to them, for free. Don't forget that the entire Internet economy was built around volunteer work We all provide content for free to Google and Facebook and they make money out of it. We not only sacrifice our privacy, but even donate our content. We are happy to do this to get free services like search engines and social media. People are increasingly telling the whole world about their habits. When you download an app on your smartphone, your smartphone asks you if you allow the app to do a,b,c,d and all of those things steal your privacy. How many people answer "no, thanks"? If you answer "no" to just one of those questions, you cannot download the app. Millions of people answer "yes" because they want the app, and their privacy is only a secondary concern. Let's be honest: the corporations and the government didn't have to fight too hard to get our privacy. The business model that succeeded on the Internet was simple: you give me a service for free, and i let you track my movements, i.e. i let you spy on my private life. The Internet economy in the age of Google and Facebook has been built around the model that services are free but the service provider (Google, Facebook, etc) makes money by selling your private life to businesses who then pay the service provider (Google, Facebook, etc) to advertise their products to you. In fact, you are the product that Google and Facebook sell to their customers. We accept this deal when we use those free services. We are eager to use those services. We are eager to be robbed of our privacy. On services like Facebook we are actually eager to tell Facebook more than Facebook asks us. We post daily news about ourselves. We work for free to make it easy for Facebook to spy on our lives. When we invented the Internet, we had no idea that your privacy would turn out to be the most valuable asset on the Internet.

It got even worse. Today we live in an age in which you look suspicious if i cannot find enough information about you on the Web. People discriminate against you if they can't find out enough about you online. Why don't you have a Facebook account? Why there is no trace of you in high school and college? Why there is no trace of you giving any talk when you worked at that company? If there is nothing about you on the Web, people wonder "What is wrong with you?" I have friends who initially refused to be on Facebook, Twitter, etc. Over the last few years they all joined at least one of these platforms. They were being discriminated because there was so little about them on the Web. After they joined, something funny happened to most of them: they realized the advantage of being "popular" and started promoting themselves as much as possible on the very social networks that they used to hate.

You better get used to it: all the connected gadgets that they will give you to install around the house or in the car will certainly collect information about you. Your private life is a gold mine. The temptation is just too big. It's like someone installating a device near a gold mine that nobody else is using. The temptation to go inside and take the gold is just too strong. An intelligent kitchen light will be sold to you at a discounted price, or even installed for free, because it can sense what you are cooking and how often, and this information can be sold to thousands of businesses and it is worth more money than the $50 of the kitchen light I know people who are already saying "there will be zero smart appliances in my house because i don't want spies in my house" but these are the same people who said they would never share their private life on social media and now they post information about themselves on all sorts of social media, and they are happy that the social platforms offer them all sorts of commercial deals.

It is easy to list cases in which people will be afraid of objects that "spy" on them. But think, for example, of the anxiety related to bureaucratic procedures. You just had to apply for a US visa. The US consulate asked you to collect a lot of documents and then scheduled a physical interview. Before and during the interview you had to live with the anxiety that you might say something wrong and be "black-listed" by the US consulate. Did you enjoy this process? Nobody does. It is a waste of time, it is humiliating, and it is even inefficient (who knows how many criminals and terrorists get the visa to the USA simply because they are good at interviews). Now imagine a world in which so many objects know so much about you that the US consulate can simply query a knowledge base and find out in a few seconds whether you can be trusted or not. The smart objects that you encounter in your daily life become the equivalent of the neighbors who know everything about you in a village. These objects can tell the US consulate "she's a good person" the same way that in the old days the villagers would say "she's a good person" because they knew you well. Smart objects will know what you buy, where you spend your spare time, where you go on vacation. Smart objects will tell the US consulate "She likes to hang out at the Lingyin monastery, she likes parks, she buys books of classic poetry, she traveled to Bhutan and she reads a lot of technical literature". The US consulate will not need an interview to grant you a ten-year visa. Yes, your privacy is gone, but you get through a bureaucratic procedure in ten seconds instead of ten days.

Attitudes towards privacy change from country to country. In January 2016 the Pew Research Center published a report titled "Privacy and Information Sharing" ( http://www.pewinternet.org/2016/01/14/privacy-and-information-sharing/ ) that studied the attitudes of people in the USA. A majority of people accept that their employer can install monitoring cameras in the workplace (in theory, to catch thieves, but obviously the cameras can be used for anything). People were split 50-50% about "loyalty cards" given away by big stores: once you get the card, the store tracks your purchases but you also get some discounts. There was only one "smart object" included in the survey and that's the one thing that people were most afraid of: the spy is coming into the house. Only 17% of people said they wouldn't tolerate any form of intrusion in their privacy. It is interesting that 4% said that they would accept all of them. This is the future: the younger generations will be more and more likely to accept all forms of intrusion. I suspect that one generation from now the numbers will have reversed.

The Facebook generation has been willingly surrendering privacy in return for celebrity. You cannot have celebrity and privacy at the same time. Choose on. Most kids choose celebrity. If you want people to "Like" what you are doing, you need to tell them what you are doing. Vanity has been winning over safety. People want to be famous and liked, not to be safe; especially young people.

I don't see how the protection of privacy can help the economy and innovation. A member of the European Commission, Andrus Ansip, said: "Today's agreement builds a strong basis to help Europe develop innovative digital services". But he didn't mention one single case. I think that agreement simply discourages startups from innovating in many fields because the business plan becomes more complicated: starting in 2018 they will have to make money by selling a product or service instead of selling information about the users, and products are much more risky to make than collecting and delivering information. If Google had to make money with its search engine, today we wouldn't have such a good search engine, and maybe we wouldn't have a company called Google at all. Unemployment of young people in Europe is very high (almost 50% in Spain, almost 40% in Italy, 25% in France). The European Union is protecting the privacy of the future generations, but the future generations will have no jobs.

In the world of humans, many things are complicated. First of all, the FasTrack card works only in California. The machine has no idea that there are thousands of regions on this planet and that each has its own system and that you are coming from another region. It still demands that you purchase a FasTrach card if you are driving through one of California's bridges. If you don't do it, then a camera takes a picture of your car when you pass in front of the sensors, and then a letter is sent to your home asking you to pay the bridge toll. To pay, you have to go online or make a phone call or drive to a "service center" and pay cash. All of these systems are more complicated than standing in line at a toll booth.

Secondly, humans are complicated. Sometimes the driver of the car is not the owner: sometimes he is someone who borrowed a friend's car. The machines have no concept of "friends". They will send the bill to the car's owner. If i sold the car to someone and the sale has not been registered yet by the state of California, i will have to pay his FasTrack fees. Sometimes humans are tourists, who sometimes don't even understand English well and cannot read the sign that explains FasTrack when they are driving over the bridge at 80 km/h. Even if your English is perfect, it would be a bit crazy to stop the car on the bridge to read the details of how FasTrack works. If you are a tourist and drive through one of the bridges of the Bay Area, you will receive a letter at home asking you to pay within so many hours. But if you are tourist, you may not see that letter for weeks. In that case you will receive another letter that tells you that now you have to pay a very expensive fine. The machines don't have any knowledge of "tourism", "far from home", "doesn't know that FasTrack exists", etc. The human world is complicated. The automatic plate-number recognition system was announced as a way to avoid the queues at the bridge tollbooths, but in reality it is just a way to minimize the costs and maximize the profits because so many drivers make a mistake or don't pay in time and then have to pay an expensive fine. If you have to deal with the bureaucracy associated with it, you will think that it is just a scam to steal money from tourists.

And here is the worst part. The machines are stupid, very stupid. I had a two-months battle with the FasTrack bureaucracy because their machines were saying that i didn't pay, but i was sure that i had paid online, and luckily i had printed out the receipt. It took two months for a human being to look into the problem. Then they called me and explained to me that i had accidentally clicked on the wrong button and paid the toll for any crossings after 6pm that day, but i had crossed at 5:59pm. The machines are so stupid that they didn't realize what any human being would have realized: that i intended to pay for the 5:59pm crossing. Because of the one-minute difference, i spent two months exchanging letters with the FasTrack bureaucracy, and those letters were mostly read by machines; and other machines, unable to understand my appeals, were sending me replies asking me to pay more and more expensive fines. Eventually the human being agreed that this was a silly case and he closed the case. I wasted more time proving my innocence than if i had stayed in a line of cars at the old bridge tollbooth of 20 years ago.

Coming back to the issue of privacy, the good news is that the bridge has a record of all the cars that passed through. If the police are chasing a car, the bridges can tell them which way the criminal drove. If a person goes missing, we can easily find out which was the last bridge that her car crossed. In the future we may have videocameras on all highways recording who is driving at what time and in which direction. It is an invasion of privacy, so maybe each driver will have to give his/her written consent for it, but i suspect that honest people will have no problem with it. The criminals are the ones who will have a problem with it: if your job is to rob banks, you don't want videocameras that can record where you were and where you are going.

It always takes time for things to "bootstrap", to switch on. We are used that things in our house just work: you press a button and they do what you want them to do. For example, you press a switch and the light goes on in your room; you press a button and the microwave oven warms up your meal; you press a button and the washing machine washes your clothes. But what happens when you press a button to turn on your computer? You have to wait. The bootstrap process on my desktop computer takes so long, that i turn on the computer and then take my shower.

Smartphones and computers like to "beep". They are not particularly intelligent, but they beep all the time, like babies. My smartphone even beeps (very loud) when i turn it off. My tablet beeps very loud when i turn it on and when i turn it off. Hardware and software engineers have a passion for making devices beep. Imagine when you have 100 "smart" devices in your homes: will they all beep when they turn on and off, when there is an update, etc? Today i have five computers in the house and each one is like a baby: there is always something wrong with the software or the hardware. Imagine the day that my house has 100 objects with a computer inside...

Smartphones and computers have become vehicles for advertising. What about the "smart" objects that we install in our homes? Will they also display advertising? Will i get an advertisement about where to buy fresh eggs whenever i open my refrigerator?

These devices are designed by hardware and software engineers, who are the worst possible people to design something to be used by everybody. Hopefully, the "smart" objects for the Internet of Things will NOT be designed by engineers.

South Korea is ahead of everybody else in the field of "smart cities". For example, the Songdo International Business District is a a 6 square kms "smart city" developed jointly with Cisco, LG CNS and ADT Caps.

There is general consensus that the potential of the technology is colossal. In the 1950s the USA built cities around cars. The USA created the first network of "freeways" in the world. The center of the city became a place for poor people while the richer families moved out, in the suburbs. The suburbs came to symbolize the great economic boom of the USA during the 1950s and 1960s. The dwelling in the suburbs was a single-family detached home with a room for each child, a garage, a backyard and a lawn. The lifestyle of the average family and the economy of the following decades were largely determined by that type of dwelling. In the next decade we will start redesigning cities around smart networks. This new approach is likely to change cities as much as the car changed cities in the 1950s. The US government started thinking about the smart city of the future in 2015 when it launched the "Smart City Challenge" ( https://www.transportation.gov/smartcity ), promoting the adoption of intelligent technologies in cities across the nation. The city that comes up with the best plan will get funding from the government to implement that plan. India followed suite with its "Smart Cities Challenge" http://www.smartcitieschallenge.in/ These programs encourage bureaucrats to look into the new technologies and imagine how those technologies can improve their cities, particularly the relationship between transportation and construction (a connection discovered in California in 1901 when Henry Huntington founded the Pacific Electric Railroad to create a network of trolley cars and a network of new suburbs around Los Angeles, thus becoming the richest man in the city). Startups look at the "smart city" in a slightly different way, bottom-up instead of top-down. They look at the smart city and see an I.T. platform into which you can plug applications/services. Urban-X is a New York incubator founded in 2016 by BMW's Mini division and Cyril Ebersweller's incubator HAX (which is based in Shenzen and San Francisco) dedicated to startups working on the city of the future.

It is not even certain that bluetooth and WiFi will be the right way to connect objects. The Internet of Things is for low-power devices, devices that have to survive a long time with a battery. LTE (Long-Term Evolution), originally introduced in Scandinavia in 2009 by TeliaSonera, uses the existing 4G wireless network. There are companies specialized in making chips for LTE communications, like Israeli-based Altair Semiconductor (acquired by Sony in 2016) and the Paris-based Sequans. In 2015 two standards were proposed: Narrow Band-LTE (NB-LTE) from Intel, Ericsson and Nokia; and Narrow-Band Cellular IoT (NB-CIoT) from Huawei and Vodafone.

Evrythng, in particular, has signed one of the biggest contracts so far: it is collaborating with Avery Dennison (a big corporation that is mostly known for its self-adhesive labels) to create "smart" clothes and shoes that can communicate with the "cloud" and therefore with smartphones. For example, you will be able to ask your shirt "where are you?" whenever you can't find it.

Google, of course, has invested in Nest, the most famous of all. The only investment company that can compete with Intel and Cisco is General Electric Ventures (not based in Silicon Valley) that has invested in OnRamp Wireless, Mocana, Quirky, APX Labs, Clearpath Robotics, and FlexGen Power Systems. In 2015 Autodesk allocated $100 million to a special fund, Forge, for Internet-of-Things startups. However, they don't necessarily invest in Silicon Valley startups. The most well-funded IoT company over the past 5 years is Sigfox, which is based in France, followed by Telogis, based in the Los Angeles area, and then Tendril in Colorado. The only major success story in Silicon Valley ("success" as in "receiving a huge amount of money") is Jasper Technologies, acquired by Cisco. Silicon Valley certainly has a lot of "early adopters". Palo Alto is now home to the first 'Internet of Things' brick-and-mortar store that specializes in IoT products. It is called b8ta and located at 516 Bryant St.

First, the hardware. Arduino Uno, that came out of Italy in 2005, has no operating system so it can only run one app at a time, but is the ideal microcontroller for "smart" objects that only do one thing. The Raspberry Pi, introduced in Britain in 2012 and originally conceived for schoolchildren by a charitable organization based around Cambridge University, is a cheap computer that fits in the space of a credit card. In 2015 TL Lim and Johnson Jeng launched a Kickstarter campaign to fund a single-board computer called Pine64, slightly bigger than a smartphone that will come equipped with the most powerful ARM processor (more powerful than the most powerful laptops) and cost just $15. And it is designed to work with the OpenHAB standard. In 2016 Next Thing, an Oakland company that had introduced the first camera powered by a Raspberry Pi, unveiled its own computer on a board, the Chip Pro, sold for $16.

Second, the interconnection. OpenHAB (Open Home Automation Bus) was designed by by Kai Kreuzer in Germany in 2010 specifically to let "smart" devices talk to each other http://www.openhab.org/ It is open source and it is written in Java. You can use it to set up a sequence of operations like "At 7 AM turn on the heating in the bathroom unless the outside temperature is already warm enough, at 7:15AM turn on the lights in the house unless there is enough daylight already." Or: "When someone rings the bell, show the security camera on the computer monitor, turn on the front lights and pause the DVD player". Or: "If everybody leaves the house and the tv is still on, send a text message". Or: "Shut off the lights of the house when the last person has left but keep the lights outside on if it is still dark and turn them on when there is enough daylight and turn them on again when it gets dark". Kreuzer now leads Eclipse IoT, an association of companies and independent engineers http://iot.eclipse.org/

Many of the open-source stories link back with Ubuntu, a Linux variant introduced in 2004 which is entirely based on free software and has become popular for personal computers and smartphones. It a descendant of Debian, invented in 1996 by a German student at Purdue University (in Indiana), Ian Murdock, who, by the way, who later moved to the Bay Area to work for SUN and died just recently in San Francisco at the young age of 42. OpenHAB is typically deployed on Ubuntu. There is a very large community of software and hardware engineers that are contributing to this story.

OpenHAB is also compatible with ROS, the Robot Operating System invented at Stanford by Andrew Ng's group in 2005 and perfected at Willow Garage. ROS is, again, open source, and provides a set of software tools to build robot applications http://www.ros.org/ ROS was the descendant of the STanford AI Robot (STAIR) that was conceived from the very beginning with "home" robotics in mind. The home assistant robot is very much a part of the Internet of Things. For example, you could have a motion-detecting device in the front porch of your house that automatically dispatches a robot to check who is outside; and then ROS recognizes the face of the person at the door, turns on the light and opens the door and asks "at what time would you like your soup?" It then recognizes the answer and programs the microwave oven to heat the soup at that time. The motion detector, the robot and the oven communicate via OpenHAB, the same way that you and I communicate in English.

The problem is how this will affect our behavior towards other humans. When we use machines, we tend to get nicer towards the machines, and less kind towards fellow humans. Think of drivers: very nice men can behave like really mean men when they are driving. The same man who is so nice when he is walking in the street can become really rude and arrogant when he is driving. When you remove the direct interaction with other humans, you literally change the way humans perceive other humans.

The second direction (exactly opposite) is the "Internet of Everything". Facebook (Internet.org), SpaceX, Google (Project Loon), Qualcomm and Virgin (OneWeb) are planning to provide global Internet connectivity to every person on the planet. Facebook's Internet.org will use drones and a satellite constellation, Google's Project Loon will use helium balloons, Qualcomm's OneWeb will use a satellite constellation (to be launched in 2018). The 2015 report titled "Measuring the Information Society Report" by the United Nations' International Telecommunication Union agency shows how far behind the developing world still is: only 7% of the households in the developing world have access to the Internet (compared with 80% in the "rich" world); but that could be the reason that the Internet of Things takes off sooner: there are countless services that don't exist and need to be created from zero, whereas the developed world already has many of those services and it is harder to replace them with interconnected things. I wouldn't be surprised if the developing countries are the first ones to adopt the Internet of Things on a massive scale, just like Africa was first to adopt smartphone payment. In 2014 a study by GSMA estimated that 52% of cellular M2M connections were in the developing world. In 2015 a study by McKinsey ("Unlocking the Potential of the Internet of Things") estimated that 40% of the value of the Internet of Things between 2015 and 2025 will be generated in developing countries. The cities of the developing world have the most pressing problems. They are polluted, congested, unhealthy. The developing world needs the Internet of Things more than the developed world to create sustainable cities. The Internet of Things can help the developing world provide health care to its citizens and to provide aid in cases of disaster emergencies. The Internet of Everything is a global village in which things create better societies than the societies built by people.