Audi Car-to-X's new gameplay, how far has the pre-installed car networking come?

The importance of data needs no further explanation. The key is how to use it.

In the past two or three years, in various interpretations of the Internet of Vehicles, similar words are often seen: when the car is connected to the Internet, it is no longer an isolated island of information, and various possibilities are available. Therefore, killer applications have become the point of competition for everyone.

From the perspective of the realization of the Internet of Vehicles, in order to make vehicles no longer "isolated islands", having the Internet function is just the beginning. To achieve "unlimited possibilities", two problems must be solved first: who to connect with and what to do after connecting . As for these two issues, every car manufacturer and service provider has its own ideas.

During last week's trip to Munich, Germany with Audi, in addition to getting close to the VR glasses and learning that I was a little 3D dizzy, I also learned about some of the progress of Audi's current Car-to-X communication project. Audi will add Car-to-X services to Audi Connect services starting this year, first on models such as A4/A5 and Q7. From the middle of this year, Audi will begin to collect various information and build a database. When the database reaches a certain scale, Audi will release the data for users for the first time at the end of the year. This service will be launched first in Germany, and the time for launching this service in China has not yet been mentioned.

3 data sharing features

In the Car-to-X technology demonstrated this time, the cloud is the first connection point and the data distribution center. Different data will have different flows. After the data is collected on the vehicle side, it is uploaded to the cloud through the in-car network and then transmitted to other Audi vehicles through the cloud. Currently, Audi is collecting three types of data and developing three functions:

1. Preset driving mode through traffic sign information

Traffic sign information is identified through the on-board camera and uploaded to the cloud. The data uploaded to the cloud needs to be verified before it can be released. The verification standard is the number of cars that uploaded this information. Only when a sufficient number of cars upload this information will it be verified and shared with other vehicles and navigation systems.

This information update method will be more meaningful for autonomous driving and the current predictive ACC system. At that time, the vehicle will preset the driving mode according to the various traffic information in the system, such as changing the speed in advance according to the speed limit information.

2. Collect and share vehicle hazard information

Currently, dangerous information includes two types of information: one is that an Audi brand car has an accident or breakdown; the other is that the surrounding environment may affect safe driving, such as slippery roads or reduced visibility due to fog and heavy rain.

Demonstration system collects vehicle fault information in real time

In the former case, when the owner activates the eCall emergency call system on the vehicle, or when the airbag is activated, the vehicle will automatically issue an accident report; for the latter, the specific impact information needs to be evaluated and verified. For example, the visibility outside the vehicle will be evaluated through the data received by the rain and light sensors on the vehicle, as well as the operating mode of the wipers. Only verified danger information will be shared with other vehicles through the cloud.

3. Sharing of roadside parking information

When the car is driving, it will use ultrasonic sensors or cameras to identify free parking spaces on the roadside and parking space occupancy information, and upload it to the cloud.

Parking space information collected in real time by the demonstration system

The sharing of roadside parking information includes two aspects. On the one hand, the system will analyze the vehicle's drive system position, steering angle, speed and other parameters to determine whether the vehicle is entering or leaving the parking space. If it is found, it will automatically send a report to the cloud server. This kind of instant information can be shared with other Audi vehicles in real time.

On the other hand, a statistical model is established based on the historically collected data and real-time data, and then the data of possible free empty parking spaces can be calculated according to the current time, which can serve as guidance information for drivers to find parking spaces.

In this process, not only the sensors on the vehicle need to be able to receive and analyze data in a timely manner, but also the data base needs to be large enough. How to cross the boundaries of brands is a question that all car companies are currently considering. Mercedes-Benz previously mentioned that Car-to-X will be open to other brands, but did not propose specific measures and implementation time.

2 functional verification methods: model car and PCC

On the cloud platform, a large amount of data can be obtained every second. Who should these data be connected to? Audi provides two different verification methods. One is to verify the technology and projects under development through model cars , and the other is to verify the technology that has been mass-produced through an Audi project called PCC (Personal Company Car) using actual vehicles as carriers.

Audi Q5 1:8 Model Car

This is the model car Audi is using for testing now. The model car is based on the Q5, with a scale of 1:8, driven by an electric motor, and can reach a maximum speed of 40 km/h. The car is equipped with two cameras - one monocular and one stereo, and ten ultrasonic sensors - five in the front, three in the back, and one on each side, which can detect lane lines, traffic signs, obstacles, parking spaces, etc. There are also acceleration sensors that can record the changing direction of the model car. Inside the car, there is a high-speed computer with a quad-core processor. The structure of this microcomputer is the same as the zFAS used in Audi's self-driving car. The data in the car can be transmitted to the cloud in the background in real time via Wi-Fi.

The model car is demonstrating its functions

At the scene, the model car demonstrated three functions respectively. In terms of the demonstration effect, the uploading and distribution of information is very real-time, and temporary "emergency" situations - obstacle avoidance staff who enter the site to test the vehicle's obstacle avoidance function can also be identified and bypassed in advance. However, this model car method is more suitable for verifying the mechanism and principle of technology, while for the investigation of information transmission speed, emergencies and other factors that have many influencing factors in the real environment, it depends on testing in the real environment. The PCC project was created for this purpose.

Audi started testing the PCC project in mid-2015. As of the middle of this year, 70 Audi managers have joined the project to test the connectivity and data sharing functions. According to Audi engineers, Germany has strict regulations on personal privacy data protection, so Audi has set up a fleet to test this function. In addition, for those who participate in the test, they can apply to terminate the test in advance when they are on vacation or when they do not want their data to be collected. Audi said that all the collected data is anonymous.

At present, these 70 vehicles have been tested for 630,000 kilometers and collected more than 400G of data. The system can collect about 500 pieces of information from each vehicle per second. After the information is collected, it will be encrypted and sent to the data background through the mobile network. According to the data type, Audi divides it into real-time data and evaluable data, and sends them to different analysis models respectively. It is reported that this back-end software for processing, storing and analyzing data was developed by Audi itself, so Audi can flexibly adjust what type of data needs to be collected and analyzed according to the situation. At present, this software will not collect all the data on the vehicle, but select the node data of 850 ECUs for collection and analysis based on the test function.

Analysis of users’ favorite radio stations based on PCC data

For Audi, this project is to evaluate and optimize the performance of these interconnection and data sharing functions in actual use. However, the post-evaluation and analysis of the data now obviously gives Audi more inspiration. According to the engineers, in addition to the conventional geographic location, driving speed and fuel consumption, the user's favorite radio station, mobile phone signal strength and the use of in-car buttons will also be collected at the same time. The use of in-car buttons can be used to optimize the center console design, but the radio station and mobile phone signal strength can be provided to the corresponding service providers to facilitate them to optimize their products.

In other words, Audi can now provide the functions that TSP has been trying so hard to provide to car companies through the PCC project. Moreover, as the scale of sample collection increases, Audi can make these data bring additional value to it. Admittedly, the change in the amount of collected data and more complex data will put forward new requirements for storage and analysis, but to some extent, this is already a successful case of car companies doing TSP themselves.