![robots]()
Over the past 25 years, the Internet has radically altered the way people communicate and share ideas and the way businesses interact with customers and clients.
For an even longer period, starting in the 1950s with the so-called Third Industrial Revolution, businesses have become more digitized. In the next few decades, a new industrial revolution will combine elements of these two trends, along with related technologies and practices, into a truly "smart" manufacturing process.
This convergence is known as the Industrial Internet, Industry 4.0 or the Industrial Internet of Things. Whatever the name, the result will profoundly affect global trade patterns, supply chains and societies. The impact will vary, presenting many opportunities for developed countries to be more disruptive in developing economies and possibly limiting the use of low-end manufacturing for quick modernization and development.
Nonetheless, this Fourth Industrial Revolution will change manufacturing, industry and society.
![RTXQPKZ]()
The Impact of the Internet
Although the Internet grew out of military applications in the 1960s, the 1990s were when its true potential for non-military uses was realized. Some remarkable changes have occurred since then. The Internet almost single-handedly destroyed many service jobs, replacing travel agents, tax accountants, bank tellers and others. Despite the dot-com bubble burst, this disruption resulted in the creation of many jobs in the technology industry.
Connecting to the digital world has revolutionized how communication occurs and how information spreads through society. People have vast amounts of information at their fingertips, and with the advent of the smartphone they can access it anytime and anywhere. Outside of the business world, the Internet has altered the relationship between state actors and non-state actors in cyberwarfare. Issues such as privacy have routinely challenged governments and corporations. Nonetheless, the biggest impact has been on consumer markets. Anyone who uses the Internet is constantly bombarded by customized ads, and users can place orders from online stores more quickly than ever.
For consumer markets, clearly the next big thing is selling Internet-connected devices such as the new Apple Watch, fitness trackers, smart appliances and other wearable technologies. Many of the same tech companies driving the expansion of consumer-focused devices are also the vanguard of applying Silicon Valley innovations and developments elsewhere, including industry and infrastructure. While the focus on consumers is obvious, the underlying potential in industry could be more profound.
![artificial intelligence robot]()
The Internet of Things
Whether the Internet of Things is used for consumer, industrial or other applications, at its heart it simply involves connecting devices to one another so they can communicate and become more efficient and effective. The Internet of Things and many of the different technologies it brings together and uses are not entirely new; the idea of the Internet of Things has existed for almost 20 years, but only now has the software and hardware developed enough for people to think about applying a wide range of its functions to the manufacturing process and industry.
There are now about two Internet-connected devices for each person on the planet, and that number could more than double within the next five years. Although most people tend to think of personal computers, tablets and smartphones as their main Internet-connected devices, these three categories together account for less than one-third of all connected devices. This share will drop in the future. To put the shift into perspective, just a decade ago personal computers accounted for more than two-thirds of all interconnected devices; now they account for less than 10 percent.
![Maker Faire Robots]()
In applying the Internet of Things toward industrial applications and supply chains, a number of associated technologies will affect industry in specific ways. Although each of them can be implemented individually — and many are already being used in some capacity — the Internet of Things is the glue that will bind them together, allowing seamless communication and analytics.
This relationship will likely form the backbone to any industrial advances in the coming years. To truly recognize the profound impact the Internet of Things has on industry, it is critical to look at the impact of some of the associated technologies.
Remote Sensors
Tiny remote sensors, continuously recording and transmitting information, would have applications in a wide range of industries, such as mining and energy. It would also have a critical multiplier effect between industries; for example, it could use information about the location of autonomous cars to more profitably sell ad space on electronic billboards.
![Rethink Robotics Baxter]()
Robotics
Industrial robotics is hardly a new concept, but today's robots are becoming increasingly more applicable to a wide range of purposes. One of the biggest manufacturing jobs overseas is assembling smaller electronic goods such as cell phones by hand. This is a repeatable process that industrial robots could very easily perform, as evidenced by Foxconn's drive to use 1 million industrial robots. The biggest benefit of using industrial robots would be replacing laborers who work long hours with robots that can work continuously except for routine maintenance.
Additive and Digital Manufacturing
Additive manufacturing, otherwise known as 3-D printing, can be thought of as an extension of robotics. Additive manufacturing can be used to print simple spare parts such as nuts and bolts near or at an end user's location. However, its biggest impact will be its ability to produce structures and delicate products that simply could not be manufactured any other way. Additive manufacturing also has the potential to produce stronger, lighter designs than traditional mold-based manufacturing because it only uses the material necessary for fabrication.
![RTR4BYA6]()
The related concept of digital manufacturing uses computer-based design processes such as modeling, simulations and visualizations to create a design concept and define the production process in a computer. With the Internet of Things in the modern factory, the designs could move quickly into production.
Big Data
A crucial component of the Internet of Things is the connection between devices and data analytics software. It is important to tame the gigantic amount of data generated by the Internet and remote sensing, since data is expected to increase more than 50-fold in the next decade.
Predictive Analytics
Predictive analytics can have a wide range of applications. It can minimize unplanned maintenance in the manufacturing process through closer monitoring of equipment health and mitigate the risk of an equipment outage by using predictive maintenance algorithms.
Automated Transportation
Google has made headlines with the success of its automated vehicles that will soon be allowed on public roads. But cars are just one mode of transportation that is being automated. Amazon has envisioned delivery drones, and there is potential to automate cargo trucks and airplanes.
![amazon drone]()
Testing for this has already begun; Nevada has approved the testing of Daimler's automated 18-wheelers on public roads. This could be a game-changer for optimizing logistics for land-based supply chains and making them "smarter."
Smart Grids
Fossil fuels have long dominated the world's energy supply. This will not change soon, but alternative sources of energy, such as solar and wind, are now critical elements of the grid in many areas. As batteries and other forms of energy storage become integrated in power grids, smart grids will become more resilient and adaptable, limiting companies' exposure to blackouts. It will also allow more seamless integration between buying and selling electricity at the local level, decentralizing the power sector.
Smart Cities
Automated transportation networks, smart grids and other related concepts, such as real-time social media locations, will make cities "smart." This can allow more predictive pre-positioning of law enforcement, more effective disaster and evacuation protocols and, of course, more efficient energy usage and traffic flows.
The Big Picture
There is a significant difference between the industrial purposes and the consumer-driven purposes of the Internet of Things. The consumer-focused aspect is driven by innovations into applications without a built-up consumer base. For example, Apple, in a sense, created the smartphone market.
![Big Data Mining]()
Industry, however, has been using many of the same principles for years. In most advanced economies — Germany, Japan, the United States and South Korea in particular — manufacturing and industry are highly digitized and automated, with frequent machine-to-machine communication. Companies have spent billions of dollars developing control systems and other connected systems.
This means that the use of the Internet of Things in industry will be more incremental than in consumer-focused applications, with evolutions happening at different times for different industries. However, regardless of the timetable, it is clear that industry is moving toward these applications.
One of the most significant outcomes of the Internet of Things in manufacturing will be the continued decentralization of the manufacturing process. The Internet has affected the dissemination of information, particularly in countries such as China, where the government tries to retain central control of information.
![Uber]()
In manufacturing, there is a clear trend toward more highly customized products for both consumers and companies. This requires flexibility in the manufacturing process and its supply and logistics chains to adapt to smaller orders and quickly shift from processing one order to another. This concept is also important because it lowers the cost barriers for developing a prototype product for companies of all sizes — even companies whose budget is a small Kickstarter campaign.
It is possible that a few decades from now, the Internet — as the backbone of modern and future industry — will be viewed with the same reverence as oil is today in its centrality to the economy.
The Developed World: Do The Rich Get Richer?
The countries set to adopt new technologies first and most effectively will see the largest benefits. This means the Industrial Internet will help create a manufacturing renaissance of sorts for low-end manufacturing in developed countries and reinforce high-end manufacturing in countries such as Germany and Japan. In short, wealthier countries are best suited to take advantage of this process.
![3d printing manufacturing]()
Even in places where the benefits will be the strongest, there will be significant disruption. Jobs will be created and destroyed. Countries like the United States and Germany have already seen most of the job destruction possible in the manufacturing sector, largely due to outsourcing. In a sense, there is not a lot to lose. However, the same is not true of the impact of the Internet of Things (and the Internet itself) on related industries. Automated vehicles, for example, will have a huge impact on the transportation of goods and services.
The Developing World: Will the PC-16 Exist?
On the opposite end of the spectrum, the industrial revolution will largely harm the developing world. More efficient, adaptive manufacturing located closer to end users is eroding the benefits of low labor costs. Interest in much of the developing world, where low-end manufacturing is prevalent, will decrease.
In countries where it is not widespread, such as most of Africa and the Middle East, the potential for a larger low-end manufacturing industry will be limited. Of course, there will always be a large market for cheap labor-produced, low-end manufactured goods — even if the Industrial Internet is widely implemented and perfected — but developing countries will not be able to attract as much investment in low-end manufacturing.
![Maker Faire Robots]()
The change in the developed world's manufacturing process will disrupt the 20th century's most consistent and widely used path for the complete modernization of a developing country.
For Japan, South Korea, Taiwan, China and even Germany, low-end manufacturing was crucial in developing their economy's industrial base quickly so they could catch up to — if not surpass — more developed economies. With cheap labor being offset by technology, fewer countries will be able to use low-end manufacturing as a growth catalyst.
China: A Unique Case
The impact of the Industrial Internet of Things on China is quite nuanced because Beijing is trying to balance two conflicting interests. On the one hand, Beijing's biggest fear is social unrest, and managing the social effects of any changes on its large migrant work force is Beijing's top priority. On the other hand, Beijing also knows that in many coastal cities labor costs are becoming exceedingly high. Many low-end manufacturing jobs cannot remain in China with labor costs in other countries, principally Southeast Asia, lagging far beyond coastal China's.
![beijing train station]()
Beijing knows it must adapt to prevent its own version of offshoring, so it has made it a high priority to elevate the sophistication of its industry to the same level as neighboring South Korea, Taiwan and Japan. China is the world's fastest growing market for industrial robots. It is quickly becoming the country with the most Internet-connected devices and is making significant strides in the adoption of additive manufacturing. However, the risk for Beijing is that these changes happen too quickly, causing short-term labor disruptions that require Beijing's swift intervention.
For China, the integration of the Internet into the manufacturing process can be thought of as an extension of the recent, rapid development of its technology, information and communications technology sectors. Although China lags behind the West in many areas of science and technology, it is becoming more competitive in engineering and high technology and could very soon rival all but the United States in those areas.
![china auto factory hainan province]()
China already has a robust startup culture in first-tier cities, whose technology corporations are performing well internationally. Weibo, the Chinese blogging site, even created a computer that bested Google and Microsoft in an artificial intelligence competition in May.
China's end goal is to leapfrog traditional evolutions in the manufacturing process, like South Korea and Japan did, and move directly to a highly automated, integrated and flexible manufacturing process that can compete in all strands of manufacturing with any country. Regardless of the short-term labor disruptions, if Beijing does not adapt, more jobs are likely to be lost, aggravating the Chinese economic slowdown and potential unemployment problems.
Cyberattacks: The Modern Arab Oil Embargo?
The biggest limitation on the Industrial Internet may not come from governments, policies or the development of the technologies, but rather the safety, integrity and security of companies' operations. Although machine-to-machine communication has been prevalent in manufacturing for decades, historically the network for that communication was closed and thus not as susceptible to external threats as networks connected to the Internet.
Some industries already are coping with the threat of cyberattacks that go beyond bringing down websites or stealing information. In March 2014, researchers at Symantec documented that attacks by the hacker group Energetic Bear had begun to focus on the energy sector, specifically targeting control systems for energy grid operators, pipeline operators and others.
![Cyber security]()
There have also been reports suggesting that hackers caused a mysterious explosion along a Turkish pipeline in 2008. In 2012, Saudi Aramco was the target of cyberattacks that infected more than 30,000 Aramco computers. System security requires constant upgrades to adapt to more innovative attacking schemes.
If the Industrial Internet is analogous to oil in its importance to industry, then a cyberattack aimed at disrupting a country's industry would be equivalent to the Arab Oil Embargo as a political weapon. Even if the physical security of systems linked to the Internet can be guaranteed, a cyberattack that affects large portions of the Internet could have notable economic effects on a country.
Of course, the impact would vary greatly from country to country. Since some countries will benefit from the Industrial Internet more than others, this also means that some countries will be more reliant on information technology. Isolating the United States from the rest of the Internet, even for a few hours, would be catastrophic for the U.S. economy, but isolating North Korea would merely be an interesting headline. Such disparities and potential impacts are undoubtedly a major issue, and countries such as the United States are incorporating cyberspace into their national security doctrines accordingly.
Join the conversation about this story »
NOW WATCH: 70 people were injured while filming this movie with 100 untamed lions
The World of Watson event drove this home. It suggested that cognitive systems have a place in almost any type of decision a person or company may be faced with, whether that involves buying a house, making an investment, developing a pharmaceutical drug, or designing a new toy.
In applying the Internet of Things toward industrial applications and supply chains, a number of associated technologies will affect industry in specific ways. Although each of them can be implemented individually — and many are already being used in some capacity — the Internet of Things is the glue that will bind them together, allowing seamless communication and analytics.
.jpg)
But if the neural network is tasked with finding a more complex feature – such as animals – in an image, it ends up generating a much more disturbing hallucination:
But that can result in software that is rather opaque. It’s difficult to know what features the software is examining, and which it’ has overlooked. For instance, asking the network to discover dumbbells in a picture of random noise reveals it thinks that a dumbbell has to have a muscular arm gripping it.
“One way to visualise what goes on is to turn the network upside down and ask it to enhance an input image in such a way as to elicit a particular interpretation,” they add. “Say you want to know what sort of image would result in ‘banana’. Start with an image full of random noise, then gradually tweak the image towards what the neural net considers a banana.”