Smart Cars: The New Game-changer in the Automotive Industry

SSR, Crystal Gacha, Ore Drop, Ten Continuous Draws, Wifi-style Hazard Warning…

Is this a new mobile game?

No, this is the new gameplay released by IM Auto, just before the Chengdu Auto Show.

For the automobile industry, this is an era of great change. The rapid response and acceleration of electric motors are gradually erasing the advantages that internal combustion engines have accumulated over the past century. The large screens in car cabins are bringing the driving experience closer to that of other digital devices. Record-breaking battery endurance and fast charging speeds are gradually making the “range anxiety” that traditional cars hold as an advantage less obvious.

As automation infiltrates gradually in cars, it is gradually changing the world’s evaluation criteria for automobiles. Car companies are using every conceivable means to create their own unique smart cars under the current trend of intelligent automation.

Incorporating game-like features and internet elements, what does a IM Auto look like?

Add a “Health Bar” to Improve Customer Loyalty

“Today’s mainstream autopilot essentially is high-level assisted driving. In the next three to five years, we will still be in a state of shared driving between humans and robots. The machine driver and human driver need to have very good communication and reminders.”

This is the response from Liu Tao, the co-CEO of IM Auto, to the recent controversy surrounding autonomous driving.

Although fully automated driving represents our ideal future image, the most basic functionality that users currently look for in a smart car is to serve as a reliable partner or assistant during driving, to help them safely navigate narrow passages or to cruise on highways to give the driver a break.

IM Auto hopes that its smart driving system can be a reliable partner, just like any other partnership, through mutual understanding and trust between the car system and the driver.

Where does trust come from? Of course, it’s from timely communication. The car system should notify the driver of the dangers it perceives as soon as possible and even let the driver know the limitations of its own ability.

To enhance the trust between the system and the driver, IM Auto has developed a Wifi-style risk warning system. By using multidimensional perception of road conditions, smart driving, and cabin information, the system dynamically predicts and displays the current level of danger the car is facing.

It’s somewhat similar to the health bar of a character in a game:

– When the vehicle system is “full of blood”, it can easily cope with the road conditions, and the driver can relax appropriately;

• When the system starts to “lose blood”, it means that the danger is increasing, and the road conditions become more complicated. At this time, the warning icon on the car’s screen starts to flash, and the beep-beep-beep warning sound and voice prompt will sound off. The DLP digital projection headlight projects a danger signal, and the seat belt also starts to vibrate. The multimodal interaction method is integrated, seeming to tell the driver: “The system is struggling with coping, always be ready to take over the vehicle.”
  

Seeing this, the author couldn’t help but exclaim that this Wi-Fi-like graded warning is truly remarkable. However, the question arises: How can dangerous situations be classified in real-time and accurately?

It should be noted that a moving vehicle faces interference factors in all directions, including lateral, longitudinal, and transverse dimensions within the cockpit, and each dimension affects the safety of the system. Even if high-precision maps and perception fusion endow the system with the ability to predict the “future,” and AI algorithms enable the system to make quick judgments, it is still a puzzling task to balance multiple dimensions of danger signals to give a rating.

However, fortunately, safety evaluations usually have “veto power.” Even if the vehicle is driving on an open highway without any interference on both sides, sudden braking of the vehicle in front may still cause the automatic cruise in the vehicle to display a signal of danger.

With this kind of visualized warning for dangerous situations, when the advanced driving assistance system is working, the driver does not need to be tense throughout the journey, which can be less tiring than driving by themselves. At the same time, the driver can also be aware of potential dangers at the first time, without waiting until the vehicle starts to brake automatically, and find the danger right in front of them in a panic when taking over the vehicle.

The trust between humans and the system is established through interactive “knowing oneself and knowing the enemy.”

About autonomous driving: Let the right “person” do the right thing

People have countless expectations for intelligent vehicles, and autonomous driving definitely accounts for a large part of it. Therefore, car companies are also very generous with their investment in autonomous driving. From partnering with technology companies to OEM full-stack self-research, each car company tries to deepen their participation in autonomous driving.

Before the Chengdu Auto Show, the IM AD intelligent driving test car of IM Auto conducted an intelligent driving challenge during the evening rush hour in Chengdu, including merging on to the highway, intelligent follow-up under congested road conditions, encountering close-range cut-ins, driving on large curvature U-shaped bends, recognizing obstacles and merging into the main road and other complex road conditions and scenarios. Although some scenes seemed to have relatively slow steering, IM AD successfully completed all challenges and achieved zero takeovers.

Smart More is confident in showing off its own IM AD capabilities during rush hour, which is closely related to its choice of automatic driving routes. In summary: Let the right “people” do the right things.

On the technical level, on the one hand, IM AD uses a hardware visual perception system architecture, through cameras, millimeter-wave radars, ultrasonic radars, and other sensors’ fusion modes to verify the system’s perception ability. In order to improve the safety redundancy of the perception system, the IM AD system provides a compatible mode for upgrading laser radar in the later stage.

According to Smart More’s preview in a previous event, what they may carry is most likely RoboSense’s (Speedtech’s Joint Creation) second-generation intelligent solid-state laser radar M1.

As the only high-performance laser radar that achieves vehicle-level mass production SOP delivery, Speedtech’s joint creation M1 has the ability to solve corner case problems of point cloud and “gaze” function that allows you to change the cross-sectional and longitudinal scanning speed according to driving scenarios, allowing the IM AD system to optimize waking perception according to different road conditions.

This may be one of Smart More’s secret weapons for providing WiFi-style danger alerts.

On the other hand, based on fully redundant hardware security and AI algorithms, real-time data generated during vehicle use is used to train and update models, driving the full-process data-driven automatic driving system iteration.

The hardware redundancy and AI algorithm work together by delegating regular scene perception to AI algorithms and delegating corner case problems to the laser radar providing safety redundancy, with both technologies playing their strengths.

Not only on the technical level, but Smart More also adopts a division of labor on automatic driving itself, cooperating with technology company Momenta, each responsible for their respective specialties: automatic driving software, algorithmic perception, and fusion aspects are handled by Momenta, which excels in software and algorithms. Planning and control, which require greater stability, security, and overall coordination, are personally taken care of by Smart More, mobilizing the suppliers of various vehicle modules to cooperate.

This division of labor logic is very similar to Smart More itself: a combination of traditional car companies and Internet companies, with complementary genes to maximize efficiency.## King Canyon + Blockchain: The Terminator of “Data Piracy”?

Data and users can be said to be the “Sword of Damocles” in the era of intelligent cars.

With data, one has a powerful weapon for intelligent technology iteration, enabling autonomous driving to grow and thrive in real-world scenarios and fill the gaps in addressing corner cases.

Having users, on the other hand, means having direct access to data’s stronghold, enabling the iteration and evolution of the system experience. Moreover, the user community becomes a solid backing for enterprises. Thus, users and data became the core driving forces of intelligent vehicles.

In the intelligent era, it’s almost unavoidable to collect user data. If data collection cannot be avoided, how can we convince users to willingly share their data? At least, it should not be through “data piracy.”

Ji believes that the plan of “Stone Canyon” is the solution. In the “Stone Canyon” system, a total of 300 million “stones” were released, and a batch of stones drops every 10 minutes. In the first phase, each batch includes 500 stones. After four years, the number of stones per batch will be halved.

Friends in the virtual currency industry might recognize this process. Yes, “Stone Canyon” is based on blockchain technology, and it works similarly to “mining.”

The “stones” are not virtual rewards; they can be used to exchange for hardware and software functionality upgrades on cars.

For example, Ji’s “angel round users” can exchange “stones” with the next-generation LIDAR fusion intelligent driving system after driving over 5,000 kilometers per year after the car delivery. After driving for three years, they can upgrade to the next-generation high-end energy battery not lower than 120 degrees.

So how can we get the stones? Ji provides two mining modes: “mileage mining” and “cultivation mining.”

By authorizing the data generated by cars to the system in daily driving, users can obtain “stones” through the “mileage mining” method.

By checking in, posting, and participating in activities on the APP, users can also earn “crystal” points. They can use the “crystal” points to draw cards, and if they draw the SSR-level reward, it’s “stone.” For gamers, this card-drawing process must be exciting.

However, compared to the usual operation of redeeming points for daily necessities or small gifts in apps, rough stones can be directly used on cars to upgrade software and hardware performance, which is quite attractive.

Finally

In the era of intelligence, we are witnessing the changes of automobiles: from traditional means of transportation becoming mobile and internet-based, and even seeing gameplay of mobile games on intelligent vehicles today, every company is trying to leave its own mark on intelligent vehicles.

However, what is important for users is not how much “addition” has been done to the car, but the more delicate and thoughtful intelligent vehicles that have been polished under these “additions”.

This article is a translation by ChatGPT of a Chinese report from 42HOW. If you have any questions about it, please email bd@42how.com.