Traditional Culture Encyclopedia - Traditional customs - How smart cockpits will evolve in the next 10 years Cars will become the third living space for human beings

How smart cockpits will evolve in the next 10 years Cars will become the third living space for human beings

Easy Car News? "Two points, one line" is the most common life norm in today's society. If the "two points" of family and company/school constitute the first and second space of people's lives, the "line" of the cockpit is the third space that deserves the name. Compared with the first and second space, the cockpit has the characteristics of mobility and more standardization, which is especially suitable for the existence of private living space. With the rapid development of automated driving technology, a high degree of automated driving can further release the driver and crew in the process of mobile travel energy, good cockpit space can help drivers and passengers to better enjoy the good life.

On the one hand, intelligent driver assistance systems need to keep an eye on the real-time status of drivers and passengers in the cockpit, and issue reminders to form interactions when necessary, and even make certain autonomous decisions based on the interactive feedback; on the other hand, drivers and passengers in the cockpit for entertainment or work activities also need a more autonomous and intelligent way of interacting in order to enhance the user experience and work efficiency.

Recently, IHS Markit launched the Intelligent Cockpit Market and Technology Development Trends Research White Paper (hereinafter referred to as the "White Paper"), and we interpreted the "White Paper" to see what the situation of intelligent cockpits will be in the next 10 years, and the future needs to be defined by us **** together.

I. Definition and application scenarios related to intelligent cockpit

What is an intelligent cockpit?

Intelligent cockpit is a silicon-based life complex that integrates intelligent and internet-connected technologies, software and hardware, and is able to realize intelligent perception and intelligent decision-making in the cockpit space through continuous learning and iteration; the difference between intelligent cockpit and mechanized cockpit and electronic cockpit lies in the fact that the intelligent cockpit is able to sense the cockpit environment more efficiently and intelligently, and is endowed with more personalities. Characteristics and embodied existence - intelligent technology and software to shape the soul of the foundation, hardware equipment to build the wisdom of the functional sensory system.

The main components of the smart cockpit include the in-vehicle infotainment system, instrument panel, head-up display (HUD), streaming rearview mirror, voice control and so on. The functions in the intelligent cockpit will be integrated into one system in the future.

The development of the intelligent cabin

The user's understanding of the value of the car has also changed from a travel tool to a "third space", and the cabin is the core carrier to realize the space shaping. At the same time, 5G, AI, big data, human-computer interaction, chip and operating system innovation will promote the intelligent cabin along the "localization-networked-intelligent" continuous upgrade. After entering the era of intelligent cockpit, the related technology is still iterating, and will eventually develop towards the stage of "the third living space".

Intelligent cockpit application scenarios

In the actual application of the intelligent cockpit, mainly including man-machine **** driving, inside and outside the joint and the "application for the king" three major application scenarios. Under the human-machine*** driving scenario, the intelligent cockpit can provide users with a certain degree of mechanical autonomous decision-making through interactive perception technology. In the internal and external joint scenario, the interaction perception of the intelligent cockpit is extended to the external environment, and the service scenario and convenience of the intelligent cockpit are extended. In the "application is king" scenario, the smart cockpit will provide users with high-quality games, video and audio entertainment services.

The key technologies of smart cockpit

Multi-modal interaction, so that the cockpit understands you better

The smart cockpit itself is a typical multi-modal interaction scenario: unlike the office and the home scenarios, the cockpit, as a person's third living space, is small in space size, relatively homogeneous in space form, and has a wide range of activities. Unlike office and home scenarios, the cockpit as the third living space of a person has a smaller space size, a relatively single spatial form, and a limited range of activities.

In the white paper, we can see that in 2021, the multimodal interaction technology that integrates voice and touch screen, emotion recognition, gesture recognition, face recognition, location positioning and so on, based on the intelligent car as a platform and comprehensive perception outside and inside the car, is becoming the trend of the intelligent cockpit.

Changan UNI-T, as the industry's first production model equipped with a multimodal voice interaction program, uses Horizon's Horizon Halo in-vehicle intelligent interaction program, which accurately determines the location of the source of voice commands by combining voice, lip movement, and other multimodal AI technologies, to avoid accidental awakening, and the cockpit can automatically lower the volume of the music when encountering a phone call.

In the future, this mode of multimodal voice interaction can make the cockpit more active and understand you better. For example, let the voice and visual, acoustic and other perceptual integration, when the user looks at the window and says "bigger", the window can be opened a little larger; when the eyes are staring at the air conditioning to say the same thing, the command becomes air conditioning open a little larger; when you are sad, say "play music", the cockpit will help you play a soothing music to soothe your emotions.... ...

Sensory system stacking, the number of sensors continues to improve

Multi-mode voice interaction, which brings us convenience, so that the cockpit is no longer a machine that can only listen to the command step by step, but can interact and have emotional communication. But it presupposes that the perception system in the car is accurate enough, which requires the role of sensors in the cockpit.

According to IHS Markit, the average number of sensors in a single car state in 2020 is 3.3, including onboard cameras, millimeter-wave radar, sign-testing sensors, and more. And by 2030, single-vehicle sensors will be in double digits, with an average of 11.3 sensors per vehicle. The most used of these in the cockpit are in-car cameras and microphones.

At the 2021 Shanghai Auto Show, newly released models such as the Xiaopeng P5, Extreme Kryptonite 001, ZhiGi L7, and Alpha S have more than one camera in the cockpit, which are mainly used for face recognition FACE-ID and driver fatigue monitoring.

Not only has the number of cameras increased, but so has the number of microphones in the cockpit. Compared to a single microphone, a multi-microphone array can better accomplish sound source localization, noise reduction, echo cancellation, and human voice separation, allowing voice interaction to expand from the main driver to the entire vehicle. As a basic hardware configuration for voice interaction, the number of microphones installed on a single vehicle will gradually increase to meet the continuous upgrade of voice interaction needs.

? Artificial Intelligence SoC - Chip arithmetic demand will continue to rise

From a technical point of view, there are at least 22 factors affecting arithmetic, each of which will have an impact on the upper layer of applications. Each influencing factor has a different impact on the algorithm, for example, the number of cameras on the full image detection is linear, other influencing factors remain unchanged, two cameras is two times the arithmetic required for one camera. For the same algorithm, it's multiple influence factors *** working together, which will lead to a significant increase in the arithmetic power requirements.

So chip makers at home and abroad are also rubbing their fists and wiping their hands to start a fierce duel in the cockpit arithmetic.

Domestically, take Horizon for example, Horizon's Journey 2 chip, with 4TOPS AI arithmetic, has been used in two models of Changan's UNI series. According to Horizon, the Journey 5, which has already completed the flow of chips, is even integrating in-vehicle human-computer interaction with autonomous driving, and completing the functions of Smart Driving and Smart Cabin on the same in-vehicle centralized computing platform. Horizon launched the Halo in-vehicle intelligent interaction solution for the smart cockpit market, the Horizon Journey 5 central computing chip for the whole vehicle intelligence, and the Halo 5.0 program based on Journey 5 will be released soon.

Three, summarize

User demand for automotive cockpit functionality will no longer be limited to the dimension of "safety", the future of the cockpit chip computing power needs to support the user demand to "active intelligence, content + service

The next ten years we will see the car will undergo radical changes, the car is no longer a simple means of transportation, it will become our travel process in the third living space, the intelligent cockpit is equivalent to our intelligent housekeeper in general, multi-mode interaction, emotional interaction, the intelligent cockpit in gradually become more intelligent, more active, this invisible thoughtful service, relying on the breakthroughs in interaction technology.

What new features will be added to the cockpit in the future, I can't say, but all new attempts and changes need to be empowered by technology as well as user needs and feedback, and we can look forward to the future of AI to take care of everything for us, and we just need to enjoy a short and cozy travel time.