How to redefine the cockpit under the transformation of intelligent automobiles?

The rapid development of intelligence and networking in recent years has pushed cars to gradually evolve into the "third space" of people's lives, and the cockpit has been redefined accordingly. At present, in the designs of many car companies and parts companies, the cockpit is no longer a cold riding space, but has begun to have anthropomorphic capabilities such as thinking, emotional interaction, and life growth, which can provide drivers and passengers with diversified services and highly emotional car experience.

However, since the smart cockpit is a complex concept involving a wide variety of subdivided technologies, the industry has not yet formed a unified definition of the smart cockpit, nor has there been a clear standard, so that companies play in different ways. In such a situation of multi-party melee, exploring a widely recognized path for the implementation of the smart cockpit has become a problem that the industry needs to solve. In this regard, many experts recently discussed the "2020 Second Internet of Vehicles and Smart Cockpit Conference" held by Gasgoo. It is generally believed that the development of smart cockpits should be user-centric and scenario-driven in order to better achieve implementation.

Technology drives cockpit transformation

The so-called "user-centric" means understanding what users need in a smart cockpit, and then taking this as a starting point to provide users with products and functions that can meet their needs. This is a phrase that many companies currently keep in mind when developing the next generation of smart cockpits.

This is indeed the general trend. With the rapid development of the four modernizations of automobiles, it is becoming increasingly difficult for automakers to retain consumers by simply selling simple functional cars. Especially as the "Z generation" gradually becomes one of the main purchasing forces in the automobile consumer market, their automobile consumption needs, concepts and paths are very different from the past. Automakers must truly stand from the user's perspective and touch their inner needs in order to gain a greater chance of winning in future market competition.

So what are the main demands for smart cockpits at this stage? At present, there are several major technical directions, the first of which is OTA technology, which is the key to giving the cockpit vitality, allowing the car to continue to grow and continuously improve user stickiness.

In the era when traditional functional cars dominate the market, the functions and performance of cars are basically determined when consumers buy them, and cannot be changed unless they change cars or spend a lot of money to change or upgrade some of the vehicle's configurations. With OTA, car companies can continue to add new functions to vehicles, allowing car owners to have a more convenient and smarter car experience, making "thousands of cars with thousands of faces" truly possible, just like today's smartphones.

Korata remote diagnosis, Image source: Korata

Moreover, OTA can also quickly fix vulnerabilities, and even predict problems before they occur, and promptly feedback risks to car owners or vehicle manufacturers to respond in advance. "For example, by monitoring the driver's driving data and the frequency of abnormalities, we can determine what kind of driving behavior will cause some components and ECUs to have problems in advance, so that we can prevent problems before they happen. Or we can carry out special customization in future car model designs for special driving behaviors of special groups of people," said Zhou Yizheng, vice president of OTA marketing at Caroda. From this point of view, in addition to helping improve driving safety and driving experience, OTA is actually a subversion of the car manufacturing model and even the business model.

The second is multi-modal human-computer interaction, which is the core of enhancing the sense of technology and emotion in the cockpit and creating an immersive experience. Multi-modal interaction can be further divided into several different technical fields such as intelligent voice, visual interaction, biometrics, and mechanical button interaction.

Among them, voice technology is the focus of the vast majority of current car companies, parts companies and related technology companies, especially intelligent voice integrated with AI technology. Whether in the L1 and L2 driver-led assisted driving stages or the high-level autonomous driving stage, it will be one of the main ways for users to issue commands and interact with system functions.

Moreover, according to Liu Qi, senior product director of iFlytek's intelligent car Feiyu OS, in addition to helping drivers complete the most basic execution actions, intelligent voice can also understand users like a partner, helping users pack everything they need during the trip with one click. For OEMs, this virtual partner also has very strong IP attributes, which can be made more vivid and three-dimensional according to the brand tone of the host, becoming an important tool for OEMs to achieve brand differentiation, such as Weilai's NOMI.

NIO NOMI, Image source: NIO

Visual interaction is mainly achieved through the use of diversified screens and cameras. Two popular technical areas in this field are HUD and DMS/OMS, and the installation rate on new cars is increasing rapidly. Among them, HUD mainly provides relevant driving assistance to the driver to avoid safety hazards caused by the driver lowering his head or diverting his eyes. It can be divided into C-HUD, W-HUD and AR-HUD. The first two represent the present, and the latter represents the future.

However, since the relevant technology is not yet mature, Deng Sunan, a senior expert in AR at Baidu's Internet of Vehicles Division, believes that W-HUD will still be the mainstream in the next few years, and will gradually move from high-end to low-end, and from optional to standard, thus entering a full-blown stage. AR-HUD is the ultimate development direction, because it can present a larger amount of information and achieve more accurate reminders. "By then, HUD will be an ecological carrier with AR navigation as the basic function. On top of this carrier, you can also have Bluetooth phones, music and other small programs to achieve diversified information display and accurate information reminders." Deng Sunan pointed out that this means that the key to the software competition of AR-HUD in the future will be AI capabilities and navigation capabilities.

An important driving force for the development of DMS/OMS is regulations. Since DMS/OMS can monitor the behavior of drivers and passengers, monitor and remind irregular driving behaviors and abnormal conditions in the cockpit, and effectively reduce the occurrence of accidents, some foreign markets have begun to implement mandatory regulations, which will be gradually implemented in the next few years. "In addition, through vision-based recognition of faces, behaviors, sitting postures, etc., a series of interactive functions can be provided for the cockpit to achieve personalized customization, such as adjusting seats and air conditioning settings according to the user's usual preferences, thereby improving the comfort of the cabin." Sun Zheng, CEO of Beijing Weidong Technology Co., Ltd., said that many domestic car companies are currently carrying out fixed-point deployment of DMS/OMS.

Eventually, these mainstream interaction methods will merge with each other to provide more comprehensive services to drivers and passengers. For example, Huang Chaojun, senior manager of human-machine engineering at Borgward and general manager of vehicle human-machine engineering development, believes that although intelligent voice is the preferred method of interaction in the cockpit in the future, visual recognition still needs to be combined. The integration of these two technologies will be the basis of multi-modal interaction. In addition, some core physical buttons on the steering wheel should also continue to be retained, such as double flash, seat door panel control, and SOS function buttons. "The era of zero physical buttons has not yet arrived, but it should be very likely to drop to about 15%." Huang Chaojun said.

Implementation should be based on the scenario first

Although the industry has basically reached a consensus on what core technologies future smart cockpits should have, in the actual implementation process, many companies still fall into the misunderstanding of "what I think is just what I think".

For example, in order to add more functions, applications are constantly added, simply pursuing an increase in quantity without any scene connection; in order to highlight the sense of technology, design elements such as large screens, multiple screens, and voice interaction are added. As for whether the interaction logic behind these functions is scientific, how many of them are really needed by users, and those that are frequently used often cannot withstand scrutiny. In this regard, relevant experts also expressed their views at this conference.

In the opinion of Jiang Aiqiang, senior technical expert on smart cockpit at Geely Automobile Research Institute, what is more important for smart cockpit at the current stage is to first improve the experience of existing functions, and then carry out technological innovation based on the scenario, including personalized function settings and the introduction of ecology, rather than simply piling up functions.

"Specifically, I think the information interconnection between mobile phones and wearable devices and vehicles, as well as security and privacy, are the top priorities for smart cockpits. In addition, there are low voice recognition rates, false wake-ups, slow responses, long OTA upgrade times, low upgrade efficiency, slow navigation responses, freezes, black screens, and Bluetooth matching functions, which are all issues that need to be addressed." Jiang Aiqiang believes that OTA is the top priority, because OTA can give cars the ability to self-evolve, and is the key to the life growth of smart cars. Without OTA, cars cannot be considered smart.