Tech Guide
AI-Powered Devices
How artificial intelligence (AI) boosts performance and improves comfort
You will encounter this abbreviation increasingly often: “AI” stands for artificial intelligence. In many professions, this technology has been around for quite some time. Now it is finding its way into everyday life. AI features turn household appliances and electronic devices into smart machines that control and optimize their own functionality. This Tech Guide shows you the advantages of this technology. Of course, your specialist dealer will be able to explain the finer details.
What does AI mean?
Intelligent machines
Artificial intelligence (AI) is one of the most hotly debated topics of our time. 96% of all Germans are aware of it*. Around two-thirds believe that jobs will be lost in the future due to AI. If not worse: the science fiction genre almost lives on the idea that intelligent machines will one day attempt world domination.
Really intelligent?
There is no real reason for such a gloomy future, however. Because even if they are called intelligent, AI systems are not intelligent in any human sense. What appears to observers to be real thinking is purely based on data and trained knowledge. Jokes and humorous answers have been programed into digital assistants like Alexa by humans to entertain their owners.
Machine learning
When people talk about AI today, manufacturers mean systems that solve a specific task on their own. One example is where a smartphone recognizes faces in a photo and automatically groups them with photos of the same people. AI programmers have trained the software to do this. Like a child learning to tell the difference between a cat and a dog, the artificial intelligence learns to detect differences. The programmers provide feedback to the software when the algorithm is incorrect and thus increase the accuracy with each new pass.
Deep Learning
Modern AI systems are able to detect complex patterns in data. They can optimize themselves, develop improved calculations from millions of inputs and become increasingly better in their field. For example, self-learning algorithms can make medical diagnoses from x-ray images. Others can detect obstacles on the road, or even learn chess or other board games without being trained by humans simply by analyzing games played by world-class human players.
Requirements
AI systems have been around for decades. There are two reasons why they are only now taking off. For the first time, there is plenty of computing power available. In 1961, a gigaflop – one billion computations per second – would have cost an unimaginable sum of 157 billion US dollars. But now, it is so cheap – every decent smartwatch has several gigaflops. The second reason: Big data. The internet, the financial industry, the energy sector, and other digital industries provide an abundance of data. The larger the collection of data used to train an AI, the better it can learn and identify patterns and connections.
AI features
The label AI is being applied to increasingly more products. Some categories of appliances and devices, such as household robotic vacuums or autonomous toys, would be unviable without machine learning. However, even televisions, home cinema surround systems and washing machines benefit from intelligent algorithms. Keep reading to find out more.
Did you know?
In 1997, a supercomputer the size of a cabinet was needed to beat the world chess champion Gary Kasparov. Today, every smartphone has many times the computing power that was used back then. Such rapid technical development means applications where artificial intelligence can be applied is now available to everyone – it’s affordable.
Digital assistants
At your service
Digital voice assistants make it easier to operate many devices in the home because they can listen and respond. “Hi Bixby, turn on the TV.” “Alexa, when will the washing machine be finished.” “OK Google, start the robot vacuum.” You only have to give a voice command. Requirement: Device and appliance manufacturers need to build in compatibility with your voice assistant. In most cases, they support only one of the common assistants. And there is one more requirement – these devices must be connected on your home network to communicate with Alexa and other assistants via the internet.
Multi-device
The most common command center in any home is usually your smartphone, a smart speaker or the voice control feature on your television. Again, what you can do with it, depends on the assistant. All devices must speak the same language, and be able to communicate with the same assistant. Your specialist dealer will be able to advise you on which ones are best. The rest is quick and easy: Link the devices in your household with the voice assistant on your smartphone, and it will obey your every command.
Did you know?
According to a recent study, 60% of all Germans have already used a digital voice assistant such as Alexa or Siri. Many do so on a smartphone or tablet (79%). However, a quarter of those surveyed also use voice-controlled smart speakers.
Housekeepers
Keep it clean
Artificial intelligence has been helping household appliances do their job for some time now. It previously was under another name: For example, washing machines had “fuzzy logic”, which would only use as much water as was required to fill the drum. Modern AI features go far beyond this, using sensors to determine the weight of the laundry inside, as well as the water. They check how cloudy the water is and dose detergent automatically.
Self-monitoring
One of the advantages of AI is how it can monitor how well an appliance is working. An appliance can determine for itself whether the filter on a dryer needs to be cleaned, or whether the water pump is clogged in a washing machine. Advanced systems also provide tips to help your resolve problems.
Analysis
Even electronic toothbrush manufacturers work with data: Thousands of brush samples taken by test subjects are used to develop an algorithm sensors data from everyone’s toothbrushes. Your toothbrush then uses this as a model to analyze your brushing behavior and give you tips on where you need to brush a little more thoroughly.
Automatic operation
Where no intervention is possible or desired, the AI can take over the control itself. Typical examples include robotic vacuum cleaners or mowers. These devices don’t like it when objects are in the way. Every obstacle makes them turn around or, in the worst case, stops them from working. Models with sensors and an AI can detect such obstacles and navigate elegantly around.
In the future
The use of machine learning and AI will only increase in the coming years. Almost all major domestic appliance manufacturers are working on this. Whereas the focus in the past was on improving existing features, new use cases are coming into the spotlight. Refrigerators with integrated voice controls are designed to recognize family members by their voice and suggest recipes tailored to suit their personal tastes.
Discolored and shrunken laundry will also become a thing of the past: Instead of setting a program manually, simply photograph the fabrics with your smartphone. The image recognition feature built into your washing machine app lets the washing machine choose the perfect settings.
Did you know?
Domestic appliances with AI features have been available since the 1990s, when the first washing machines with “fuzzy logic” came onto the market. The “fuzzy logic” allowed appliances to automatically adjust water consumption to match the weight and absorbency of the laundry – this was an early form of artificial intelligence.
Cars
Digital co-pilots
Many drivers come into contact with AI regularly, without even suspecting it: Most often when their smartphone automatically calculates how long the journey to work will likely take. Thanks to artificial intelligence, this task is done without even being requested, and immediately suggests a route that avoids the morning traffic jam.
Progress
Park and emergency braking systems use clever algorithms to assist. In a few years, they will take over the wheel completely. In an autonomous car, almost nothing runs without AI. It analyzes data from radar and ultrasonic sensors to perceive its surroundings. The AI decides in a split second whether it is the right time to steer or brake.
Safety
Traffic experts estimate that the number of dangerous accidents will subsequently decrease. No wonder: around 90% of all accidents across Europe are due to human error, according to studies.
Did you know?
The development of autonomous cars is happening in stages. Today’s cars use assistance systems (level 1), or temporarily take control of the wheel (level 2). At level 3, the drivers can take their eyes off the road for an extended time. At level 4, while the driver must be fit to drive, the car drives itself. At Level 5, there is no driver – the car only transports passengers.
Photo & video
Better than the camera itself
AI systems are particularly good at image and pattern recognition. With self-learning algorithms, they can analyze photos or videos well, running through a multi-stage process.
Image recognition
First, the system identifies light and dark pixels. Then it looks for edges and simple shapes in the image. The structures that are recognized become increasingly complex until the algorithm has learned which patterns indicate a face. This knowledge is then built into a camera or software to process images.
Optimization
Smartphones can blur the background behind the subject in a portrait photo, which would otherwise need a large format camera and a zoom lens. Or they extract the faintest details from night shots that would need to be exposed for minutes without the AI function. In televisions, the technology converts low-resolution video signals to match your screen’s resolution.
Did you know?
One typical application of pattern matching is Google’s image search function. The search engine analyzes an uploaded photo and suggests matching or similar photos. The same technology can be used to recognize and anonymize faces – the people on Google’s StreetView are automatically rendered unrecognizable.
Audio
Always the best sound
In modern audio systems, AI features improve the sound. Artificial intelligence continuously analyzes the audio signal and dynamically adjusts the frequency response, volume and other settings.
On the road
This is particularly practical in noisy or changing conditions, like when you commute to work on public transport. Active noise suppression in headphones isolates you from your environment. But it doesn’t let you hear announcements on the platform. Headphones with AI detect your situation by the noise level, and then change the noise cancelling settings or activate an outside microphone – so that ambient noise can reach your ears when necessary.
At home
A rich, room-filling television sound sometimes can be achieved only with the loss of speech intelligibility. Action-packed blockbusters drown out voices. TVs and soundbars with traditional dialog enhancement only partially solve the problem. They lower the bass and treble, but make the mid tones louder. The result is a flat sound. Changing audio settings manually each time is annoying. When AI is in control of the sound, the device notices when you are watching Star Wars, a car race or the news and adjusts the audio settings to match. Some TVs use smart algorithms to convert dual-channel stereo broadcasts into virtual surround sound.
In the home cinema
AI features have another practical advantage: AV systems are easier to operate. Large AV receivers for home cinema systems come with dozens of surround settings, but in the past, you wouldn’t use most of them because it’s more convenient to keep your receiver in its default mode. With artificial intelligence, the receiver automatically adapts to the program you are watching and sets the stage appropriately.
Did you know?
Artificial intelligence can also compose: A smartphone manufacturer finished Schubert’s unfinished Symphony no. 8 with intelligent algorithms. They analyzed the timbre, pitch and measure of the first two movements in order to calculate the rest. It premiered in London in early 2019.