State of AI

Ubiquitous Intelligence

A few years ago, artificial intelligence (AI) was more promise than reality, but today it’s part of everyday life. For instance, voice assistants on our smartphones respond to verbal questions. They can find inexpensive flights or dim the lights in the living room. In another example, given thousands of vacation photos, AI software easily picks out the faces of our friends and family. And autonomous vehicles – albeit still in test mode – are driving through some cities and on many highways making decisions on their own.

All of this sounds promising, and there is indeed good reason to be hopeful. Rather than simply executing a list of commands by rote, systems based on artificial intelligence use so-called neural networks modeled on the human brain to learn from experiences and improve in real time. The more we shop, watch movies, or whatever, the better an AI product’s recommendations become. For some people, this accuracy may feel a bit disconcerting, because they may feel as though someone is watching them. On the other hand, people naturally also appreciate an assistant when they are unsure what to do, or when they are in a hurry. For example, “Watson,” the cloud-based cognitive system developed by IBM and trained on huge quantities of data, is already helping doctors make more accurate diagnoses, improving insurance companies’ risk estimates, and giving meteorologists the ability to make increasingly precise weather forecasts.

What effects will this potentially revolutionary technology have on employment, societies and companies? In what business sectors will intelligent machines displace human beings in the coming years and decades – and where might AI give rise to new professions? How can companies use AI to tap into new fields of business and offer their customers new services that promise to increase productivity and efficiency? These are complex questions, and the digital assistant in our smartphones doubtless has no answer to them today. But one thing is clear: AI will bring about lasting changes.

Interview

“Software can be very intelligent. But it cannot replace people.”

Rolf-Dieter Heuer is confident that clever machines can help us  –  in research, industry and daily life. In an interview, the former director general of CERN and long-standing president of the German Physics Association (DPG) explains why we will still need intelligent people as well.

Artificial intelligence is a term that inspires hope and raises concern. What does it mean for you as a physicist?

 

Artificial intelligence used to mean simply intelligent software –  neural networks that were increasingly good at identifying patterns thanks to growing volumes of data fed to them by people. Software was written in such a way that it could continuously advance itself with help from people. At first glance, this is still how it works today, but a second look reveals a difference: People are giving the software an immense quantity of data and writing it in a way that makes it is intelligent enough to comb through the data looking for patterns on its own and to select just the combination people need.

 

Do you have an example?

 

Let’s take the control system of the particle accelerator at CERN. It is equipped with millions of sensors to monitor the proper functioning of everything. Various sensors are called on simultaneously as patterns are recognized. The question is not: What is this particular sensor doing?, but rather: Which situation results? With intelligent software that combines the knowledge acquired from all sensors, we can quickly say: We’ll ignore that one sensor. But if the neighboring sensor displays the same discrepancy, the program has to be intelligent enough to determine: Now we have to do something.

Time to market is reduced and the products are better if you use smarter programs. But: Smart programs are written by smart people.

Prof. Dr. Rolf-Dieter Heuer

So people no longer play a role?

 

They do. But these decisions need to be made in such short time that one cannot depend on people alone. People are too slow. A machine is certainly faster –  but will it one day be smarter? Despite all the artificial intelligence, the value of human experience should not be disregarded. I cannot imagine that machines will one day override us. The same is true with autonomous driving: Software can be very intelligent, but it cannot replace people.

 

So, what opportunities does AI open up for us?

 

What we are talking about is a cycle: In order to make advances, basic research needs ever more intelligent programs. Such new technologies naturally accelerate applied research and thereby drive industrial applications. With appropriately skilled people, this then advances the methodology, which in turn impacts basic research. This cycle is the most important thing we have. Because it consistently produces something that is disruptive, a true novelty –  and it always happens at an unplanned location. Time to market is reduced and the products are better if you use smarter programs. But: Smart programs are written by smart people.

Challenge

From Industry 4.0 to Smart Cities

AI is shaping our lives not just in B2C fields, but also in sections of the economy, such as infrastructures, industry and energy distribution – yet in ways that are only indirectly noticeable. 

Take Industry 4.0, for instance. If production is not just controlled by software but is embedded in a connected and adaptive environment, it can be constantly adapted to new conditions and thereby optimized. Control software can identify trends and patterns in the massive amounts of data generated by a factory and help make production more efficient. One way it can do this is through predictive maintenance, which uses the data generated by machines to schedule service operations that minimize production downtime. Thanks to this overview in real time, AI assistants can also support workers by constantly providing them with up-to-date information. Of course, the more connected an installation is, the more important cybersecurity becomes. And AI helps here too. If an attack is preceded by a pattern of suspicious signals, monitoring software may be able to repel it more quickly and successfully than a person can.

Energy distribution is another major area in which AI is already providing substantial benefits by helping to maintain a smooth flow of power despite the increasing use of distributed power generation from renewable sources. The healthcare system too is benefitting because AI is opening the door to analytics that allow tailor-made treatments. And considering that, according to the United Nations, 70 percent of the people on earth will live in cities by 2050, densely populated areas will naturally use AI to maximize quality of life while minimizing the use of resources. To accomplish this, AI systems will analyze data sources from transportation to energy demand and climate. The resulting information will allow cities to develop mobility platforms that allow people to easily get from point A to B without a car of their own. Fewer vehicles will mean more space and therefore a higher quality of life.

Infographic

 

Siemens – a Technological Leader

An Operating System for Industry

Siemens plays a leading role in all of these areas. In the 1990s, Siemens researchers were among the first to put artificial neural networks to use for innovative solutions such as the optimization of energy-intensive steel mills. Today, there are more than 200 experts at Siemens working on data analytics and neural networks. They are using deep learning methods that work with many thousands of simulated neurons. “Our experts are helping to create AI applications for industry, smart cities, trains, medical engineering, wind farms and energy distribution,” says Siemens Chief Technology Officer Roland Busch. “We are frontrunner in industrial AI.”

 

MindSphere, Siemens’ open, cloud-based IoT operating system – what Android is to smartphones, MindSphere is to industry – is already being used by almost a million devices around the world, with 1.25 million link-ups expected by late 2018. MindSphere connects industrial plants with the cloud. There, data can be securely collected, evaluated and used by a large variety of apps for various needs. All of this allows plant operators to, for instance, engage in predictive maintenance, energy data management and resource optimization.

 

AI-based systems from Siemens are also helping to autonomously optimize gas turbines. For instance, by analyzing a turbine’s operations, AI software can learn how to generate more power while emitting fewer nitrogen oxides. By the same token, such software is also improving the performance of wind turbines. In predictive maintenance for trains, AI programs are helping to enable shorter servicing times. Even the difficult configuration of switches for railway stations is supported by AI. “Our system independently generates an architecture for hardware and software that demonstrably complies with all safety requirements,” says Siemens AI expert Michael May. And in the field of medical technology, AI systems from Siemens are helping doctors to evaluate thousands of X-rays.

We are frontrunner in industrial Artificial Intelligence.

Roland Busch, Siemens Chief Technology Officer

Prospects

A Partner or a Competitor?

These examples illustrate the extent to which AI has already become part of our lives. What’s next? In the near future, there will be autonomous vehicles, factories that respond in real time to changing customer demands, surgical robots and intelligent digital assistants. The result of these developments is likely to by improved quality of life for millions of people. We can expect to see more sensibly managed traffic systems; increasing use of evidence-based healthcare, with AI supporting doctors with diagnoses; and energy distribution systems that automatically learn to keep supply in sync with demand. Furthermore, when in doubt about our actions, AI will help us to make decisions. By analyzing huge quantities of data, it will provide ever-more precise forecasts regarding complex systems such as the climate – and thus help us to better estimate the consequences of our behavior.

 

It’s true, however, that AI holds the potential of eliminating countless jobs now performed by human beings. With the advent of autonomously driving trucks, for example, the logistics industry will need fewer drivers. Government, society and business must be prepared for such changes. We will therefore need road maps to guide us through tomorrow’s digital infrastructures, targeted training programs and a legal infrastructure that not only paves the way for AI, but also keeps it on the right track.

 

At the same time, most experts agree that robots are a long way from replacing human beings. More likely, they say, is a scenario in which robots and AI systems will become our partners – as voice assistants are today. They will support us and, in so doing, give us the freedom to accomplish new things and perform work that requires more creativity. “Machines will always think differently from people, so I see them not as threatening but as complementary,” says Michael May. “In many of these areas, AI is like something that boosts intelligence without making people themselves redundant.”

Inside Siemens' Labs

Artificial Intelligence at Siemens

Picture Credits: from top: 1. and 3. picture getty images, 7. Empire City by Ole Scheeren, 9. shutterstock, 10. mauritius images