Last week, the Ministry of Industry and Information Technology pointed out that with the development of information technology, data has become a national basic strategic resource, and has become a powerful engine for stimulating technological progress and industrial development, and will certainly lead to the arrival of the digital economy era. As you can see, big data applications are the next chain to compete for layout.
However, data collection is the foundation of platform applications, no matter which field. This is especially true in the field of car networking, where giants and car companies are now involved. However, because each data collection specification is different, and there is no unified standard for the definition of data items, the reality is that many collected data cannot directly generate value, and the later modification period is long, the data maintenance cost is high, and resource waste is caused. Nor can it meet the needs of users. So the industry has a voice, do you need to develop a unified industry data collection standard?
How to define data standards?
On December 2, the former car network big data service operator Rainbow Wireless, together with SAIC, GAC, Jiangling, Chery, Zotye, Bosch, Desaixiwei and other 22 auto manufacturers and car network hardware and software service providers, jointly launched the first A pre-loaded car network data collection standard.
On the morning of the official release of the standard, Che Yun was invited to participate in the standard seminar, including SAIC, GAC, Dongfeng Nissan, Changan Automobile, Jiangling Motors, Cheetah Motor, PICC, Pacific Insurance, China Life Insurance, China United Property Insurance, More than 70 corporate stakeholders, including Dadi Property Insurance, Relief Insurance, Swiss Re, China Automotive Technology Research Center, and Peking University Business Intelligence Research Center, discussed the data standards and application values.
Specifically, the standard is to standardize the content and frequency of data collection through a file, to uniformly define the coding and description on the data, to uniformly define the unit and accuracy on the data, and to verify the model based on the data cycle of 100,000 vehicles for two years. And application, based on the opinions of more than 30 car manufacturers and insurance companies, summed up 135 data items and carried out unified coding.
First, the acquisition criteria include data accuracy and periodic frequency. Among them, Rainbow Wireless divides the data collection period into two categories, one is periodic acquisition, and is collected according to a certain cycle during the running process; the second actual acquisition is only when An event that is collected at the time of a specific event, such as the state of the vehicle when it is ignited.
Secondly, in terms of data quality, we know that the operation and service basis of big data lies in establishing a business model of related data, and the most fundamental data service model is data quality. Data quality includes two aspects, one is the data collection category, and the other is the data precision, so the data used in different fields and scenarios have different quality requirements.
For example, many insurance companies are currently pricing UBI products, and most insurance companies are based on the model of driving safety. The data here includes high-frequency brakes and loops. These behaviors are The ephemeral, data acquisition requires at least one second of frequency to accurately analyze the above driving behavior. If there is no strict definition of the specification of T-BOX data collection during the construction of the previous car network, then there will be insufficient precision, which means that the data lacks guidance for the pricing of insurance products.
Huang Liang, founder and CEO of Rainbow Wireless, said: "Talking about big data, there is no focus on disengaging applications, and the application model must be based on data that matches its business model. From the source, it also relies on high quality. The data."
Che Yunyu believes that the core of the collection standard of the pre-installed car networking data developed by Rainbow Wireless lies in the data quality. Based on the model, the required data items such as vehicle speed, latitude and longitude, continuous fuel consumption, etc. are deduced, and then the data can be achieved in multiple dimensions. Accuracy requirements, such as GDP up to 8 digits or vehicle speed in units of per second, redefine the standardization of data.
Rainbow Wireless ç¥ Super explains how to judge whether it is frequently changed, Rainbow Wireless needs to collect
1. The user's accuracy and latitude, while using the low frequency band GPS acquisition frequency;
2. Collect the speed of the vehicle and the speed of the steering wheel to determine whether to turn frequently;
3. Collect the acceleration of the vehicle to determine whether the vehicle is driving at a high speed or the like.
On the other hand, Rainbow Wireless has the advantage of large data collection for front-mounted car networking, that has reached data operation cooperation with 18 vehicle manufacturers, and is also engaged in technology and product docking with five auto manufacturers such as SAIC and Zotye. In the next few months, they will be online. In addition, the first phase of the pilot work with Shanghai Anjixing has been completed.
How to apply the data model?
Based on the integration of high-precision data and building application models, Rainbow Wireless will mainly apply this standard in two aspects, one is the cube plan based on insurance innovation, and the other is data intelligent service based on the optimization of OEM business.
On June 2 this year, Rainbow Wireless positioned the company's strategy as a service operator for pre-installed vehicle big data. The two lines of business are based on the insurance innovation cube program and the innovative and optimized depot data intelligence service. It can be seen that either The company's positioning or business line selection is based on the data depth mining and application. The industry-acquired data collection standards released this time are also in the overall strategic layout.
It is reported that in the field of insurance innovation, Rainbow Wireless has built a driving safety rating model, UBI pricing model, collision recognition model, and claims anti-fraud model. A strategy of individualized differentiation based on value behaviors, matching risk value with actual risks, such as a series of applications such as active reporting and anti-fraud.
In the intelligent application of vehicle data, Rainbow Wireless superimposes existing CIM and DMS data, and builds spare parts quality analysis model, vehicle performance evaluation model, fault warning model and customer life cycle analysis model. Trying to provide a basis for decision support for the entire R&D, production and after-sales of the car manufacturer, reducing R&D costs.
It is worth noting that in order to get more resources from the vehicle manufacturers and related companies in the car networking industry chain, Huang Liang puts aside: "Where the data collected by this standard is used, Rainbow is willing to bear certain data traffic costs."
Double Iron Hotplate,Double Burner Electric Hotplate,220V Hotplate,Double Burner Hot Plate
Shaoxing Haoda Electrical Appliance Co.,Ltd , https://www.zjhaoda.com