Artificial Intelligence goes Twinning – making analytics come alive
Jivka Ovtcharova, February 2019
We live in exciting times. Artificial Intelligence (AI) is fundamentally transforming the world with all its facets. Even if it has been around for a while in science and technology, aiming to create machines with some form of human intelligence, over time AI has become an inseparable part of our daily lives, personal as well as professional. Online searching engines, financial transactions, digital personal assistants, image and voice recognition are just a small selection of the most popular applications of AI today.
Behavior analytics follows on data analytics
Unlike the current mainstream, i.e. in the media, communication and financial sectors just dealing with data, the AI of tomorrow will be more about perceptible presence, based on the fusion of contents. Based on current market research and own insights the main trends for 2019 clearly indicate the transition of data analytics towards augmented analytics, called also an analytics of behavior.
Due to the rapid development of human-centered platforms, social networks and AI applications in daily live the AI refocuses the way of consideration from the “object” (machine, computer) to the “subject” (human being). People’s roles are getting more and more important. Furthermore, people are developing an “AI mindset” in direct communication with other people, but also with machines and computers. The term still sounds like a buzzword, but we’re already linking our personal attitude to AI, the way we think, act or feel. It is also about our human skills, competences and experiences in dealing with AI, which belong together and are reciprocal.
To understand the matter of which a human being, a physical or an artificial thing consists and how they behave requires the analysis in context. A mash up of AI and business technologies, such as Internet of Things (IoT), Cyber-Physical Systems and Automation will begin to make a significant impact on AI market in 2019. Edge and Fog Computing, Mesh Wi-Fi, Beacons, Blockchain, 5G, Immersive Experience and Quantum Computing – the list of technological trendsetters for future AI is long. However, it is not the technologies that drive changes, but the acceptance, the way people perceive and use technology. Thus, it is already possible today to recognize relationships and interactions that can be symbolically captured in a four-dimensional “all-in-one” behavior space characterized by intelligence, immersion, infrastructure and real-time capability.
Artificial Intelligence expands human intelligence
The boundary between “online and offline” disappears influencing the way people imagine reality in space and in time. Material and immaterial worlds merge. Real-time applications, supported by realistic visualization technologies make invisible phenomena visible and understandable to humans in order, for example, to realize new product features and functions. Numerous examples of the use of AI, such as intelligent voice assistants and in-store beacons, show that the mass markets will rely on technologies that are available, affordable and intuitive to use, enhancing experiences of customers and evaluating their data in context. At the same time, the immersion level of people’s perceptions is increasing as both Bluetooth headsets and speakers and other peripherals such as VR/AR headsets, smart watches and glasses are becoming more prevalent.
Furthermore, it is also about digital business, education and qualification. As a result, prejudice, traditions and physical constraints (including availability and local presence) become less important. Networked thinking skills, with a sense for the big picture, are in demand as never before. Life is analogue, but communication is getting more and more digital. This trend offers unimaginable potential for new professions, qualifications and business models.
Today, companies from a very wide range of industries express their extreme commitment to bringing AI into the core of their business and plans for growth. The most important thing is to get rid of the stereotypes that are outdated and to look forward with the courage to disrupt. Due to the transition of AI to more behavioral analytics not just Silicon Valley companies are on the trend anymore. Indeed, companies driven by engineering innovations, i.e. Digital Twins, take the chances to create new AI markets making analytics to come alive, driving actions and delivering value.
Digital Twins make analytic come alive
The notion of a Digital Twin is now being widely adopted. It is rapidly becoming the technology of choice for virtualizing the physical and digital worlds. As versatile and powerful as AI may be the original purpose of the Digital Twin remains unchanged: to enable people to perceive realities, study problems more easily, get to the point, understand, decide and proceed pragmatically and rapidly. Digital Twins strengthen the “human front-end” of all we do, making it more dynamic, faster learning, and also highly interactive. Furthermore, they offer excellent opportunities to investigate the unexpected and discover the very best solutions – true to the motto “It is not the technology that changes the world but the way people use it”.
But, what is the Digital Twin exactly? Similar to the evolution of AI, the notion of Digital Twin began around 60 years ago, in the pioneering era of the space exploration. At that time, the US National Aeronautics and Space Administration (NASA) was grappling with the challenge of designing objects that could travel so far away they would be beyond the human ability to see, monitor or modify them directly. NASA’s innovation was the “digital twin” of a physical system – a comprehensive digital double which people could use to operate, simulate and analyze an underlying system led by physics.
Fundamentally, the Digital Twin is a virtual representation, an embodiment of an asset of any type, material or non-material – including everything from power turbines, plants and buildings to services and maintenance. The Digital Twin is described by the structure and behavior of connected “things” generating real-time data. That data is stored usually in the cloud or edge and analyzed with relation to the running environment around it. It is then presented to users from different perspectives and in a variety of roles, so they can remotely understand its status, its history, its needs, and interact with it to do their jobs.
The interfaces to external systems and validation environments with consideration of all relevant resources and processes ensure high-level connectivity and are the key for success. For example, using the Digital Twin, it is possible to validate operational concepts for production systems in real-time, for manual as well as automatic operations, and for configuration via intuitive man-machine interfaces (e.g. web surface, haptic interaction devices). This makes it easier to take decisions based on up-to-date, transparent information. Thus, by merging real and virtual environments, intelligent commissioning of production can be used to generate forecasts based on real-time data from the shop floor.
Engineering of Digital Twins
How Digital Twins work? A ready-to-use solution at the Industry 4.0 Collaboration Lab at the Karlsruhe Institute of Technology (KIT), Germany, offers a good example. The use case of a Digital Twin of a milling machine used for process optimization and networking in virtual reality, while taking account of resource flows, demonstrates the practical advantages of the proposed solution by increasing productivity more than 20 percent. Systems which make it easier to gain a unique, deep knowledge of assets and their behaviors throughout the life cycle will pave the road to achieving new levels of optimization and business transformation. For example, we want the physical build to return data to its Digital Twin through sensors so the Digital Twin contains all the behavioral information we would have if we inspected the physical build itself.
According to the German Association for Information Technology, Telecommunications and New Media (BITKOM), Digital Twins in manufacturing industry will have a combined economic potential of more than € 78 bn by 2025. However, this potential can only be achieved if Digital Twins are implemented in a comprehensive and self-optimizing manner which enables them to adapt to future changes. Current studies show that mature and widespread implementation has not yet taken place. There are three main weak points to be discussed: model semantics of Digital Twins is mostly geometry driven, data analytics is aligned but not embedded into the model and simulation and user-interaction take place offline, due to the huge model complexity and the lack of computational power. To overcome current limitations, the following three conditions must be fulfilled: model semantics must be usage driven and adaptive for changes, analytics should be embedded into behavior and working in runtime, and the implementation should go along with experiments and experiences.
Implementing Digital Twins demands that we put real problems “into the sandpit“ of business units, think, try out, create “all-in-one“, apply emerging AI technologies playfully and quickly, test new solutions in runtime to gain experience fast and transform knowledge into actions and skills. The time to act is now! We invite you to join us in establishing “AI Twinning” as a trademark.