As a branch of computer science, artificial intelligence (AI) deals with the automation of intelligent behavior and machine learning. The attempt is made to emulate certain decision-making structures of humans, so that a computer or a machine is able to solve a problem relatively independently. However, AI can equally refer to an imitated intelligence that simulates intelligent behavior based on algorithms. This type of AI is used in computer games, for example.
Data forms the interface between IoT and AI
Connected things are delivering data on an unprecedented scale. This data plays an important role in two ways. On the one hand, the collection of data serves to make certain things smarter than they were before in their non-networked state. For example, Amazon’s Dash button, which uses a sensor to sense whether it is being pressed and then translates that information into an order, which is then forwarded via the Internet. In one form or another, all IoT devices perform real-time analytics of this kind and sense their surroundings. On the other hand, data and data analytics play a central role in the success of the IoT from the perspective of the companies that make connected products. Companies need to learn how their products are actually used and how they can be improved by analyzing usage data. For example, what happens when a child presses the Dash button 23x in a row while playing? In contrast to real-time analytics, this is also referred to as post-event analytics. Through the use of deep learning or machine learning, patterns can be identified and analyzed that help to better understand the use of IoT products and to better control the processes linked to them.
AI connects things intelligently
Many connected things, such as a smart robot vacuum cleaner, function as stand-alone units, but they can be integrated into a broader ecosystem such as a smart home via the cloud. Only when these units are capable of learning from their daily use and drawing intelligent conclusions do they really become smart things. This is also referred to as adaptive intelligence. In the case of a vacuum cleaner, for example, this could mean that it gradually gets to know the apartment and can be sent to the kitchen to clean up after a certain point. In this context, the integration of AI also increases the security of IoT devices. This is because monitoring and pattern recognition can better distinguish between normal use and external attack.
The IoT needs Artificial Intelligence
The IoT is primarily about collecting, analyzing and using data on everyday objects and things. The better the data can be analyzed and translated into meaningful use, the more likely companies and end users will be to offer or adopt IoT devices. AI offers the opportunity to analyze the data generated in the IoT environment better than ever before. Machine learning algorithms, in particular, can be used to better integrate IoT solutions into given contexts accordingly, both because it can improve real-time analytics, but also because it can be used to derive recurring patterns in post-event processing. A recent PwC study also ended with the conclusion: “The IoT needs smart machines. So there is a need for AI.” Conversely, this also means that only when the IoT and AI are consistently thought of together can intelligent solutions be developed that generate real, meaningful added value from the data.
Only the combination of IoT structures with artificial intelligence enables the creation of intelligent solutions that deliver real added value from the data obtained.
Many areas of application are conceivable or already in use:
- Local and long-distance traffic control (for example, self-driving subways).
- Smart home (e.g., robot vacuum cleaners or refrigerators), smart transportation or fabric
- Security (including cloud-based security solutions)
- Customer management and marketing