The world’s first mass-produced vehicle model equipped with three laser radars, a complete automotive solution coming out
On April 15, the BAIC ARCFOX Alpha HI, equipped with Huawei’s automatic driving technology, conducted a public test ride in Shanghai, which is also the first public test ride of Huawei’s automatic driving technology in the world. Xu Zhijun, Huawei’s rotating chairman, said, “the R & D team told me that Huawei’s automatic driving can achieve 1000 kilometers without intervention in the urban area.
In addition, at the 2021 Shanghai auto show, the “Huawei inside” cooperation model provides the passengers with an autonomous driving experience in urban areas with dense vehicles.
According to the official introduction, the HI model is the world’s first mass-produced model equipped with three laser radars. It has Huawei’s highest level of automatic driving and covers all scenic spots in urban areas, high-speed roads, and parking lots.
SAE International divides automatic driving technology into five levels, from L0 to L5
L0： No Automation
There is no automatic driving. Needless to say, the driver can’t be distracted at all.
L1： Driving Aid
The vehicle is still controlled by the driver and has some primary driving AIDS. In fact, this is very common now. For example, constant speed cruise and automatic parking. However, they can only play an auxiliary role, the main driving force is still the driver.
L2 ：Partial Automation
The vehicle has many automatic driving functions, but the driver still needs to take the lead in driving. The L2 level is currently an automatic driving system equipped by most new cars in the market. For example, new energy cars. The car officially claims to have the most complete Adigo automatic driving services, with functions such as high-speed automatic driving assistance, congestion automatic driving assistance, automatic parking, and automatic emergency braking.
L3 ：Conditional Automation
On this level, the driver can “release” himself, but he still needs to keep his nerves tight and observe all the time. He has to be ready to take over the driving in case of emergency. Therefore, L3 is also called conditional automation. However, the importance of the driver is decreasing.
L4： Highly Automated
On this level, the vehicle almost replaces the driver. Playing the role of a teacher, the driver can either have a rest or take over the vehicle alternatively.
L5： Fully Automated
L5 level is more ideal, which means that the vehicle has completely replaced the driver, and there is no need to worry about any weather or geographical factors. In the future, the car will be transformed from a car to a cabin. Under any condition, the intelligent computer can control the car. Of course, the driver can also choose to operate it.
In California, more than 60 unmanned car companies around the world have been approved for road testing.
In the “autonomous driving takeover report” released by the California Vehicle Administration (DMV) in 2019, it is stated that Baidu has surpassed Waymo in ranking first!
Here comes the question: how to measure their strength?
We need to introduce a concept: the number of times to take over
In the process of automatic driving, if there are problems that cannot be solved by automatic driving, human safety officers will take over the control, and the number of times that human beings take over the control represents this indicator.
Baidu’s team in the United States ranks first, driving an average of nearly 30000 kilometers before it needs to be taken over by human beings.
Waymo, Google’s driverless car company, ranks second, driving an average of 21300 kilometers, before a takeover.
Cruise, GM’s driverless car company, ranks third.
With the development of electric vehicles, the improvement of sensor and chips technology, the breakthrough of machine learning algorithm, the popularization of cloud computing, many technologies and traditional automobile companies around the world have begun to take steps, looking forward to standing out.
At present, the major unmanned vehicle companies are speeding up technology research and development, trying to achieve mass production.
The more accurate annotation is, the better algorithm performance will be. Any tiny error during a driving experience may lead to dreadful results. Nowadays, people are more and more concerned about the driving safety issue as several self-driving automobile accidents happened.
With the tremendous amount of training data and the high accuracy requirement, a high-quality data annotation service is crucial to guarantee autonomous vehicles are safe for the public.
Back to Tesla, this company uses cameras for visual detection, each car is equipped with 8 surround cameras. There are more than 750,000 Tesla cars around the world. If a Tesla user drives one hour a day on average, 180 million hours of video record can be generated per month.
Tesla Autopilot project has included 300 engineers plus more than 500 skilled data annotators. The company plans to enlarge the data annotation team to 1,000 people in order to support the data process. In an interview, Elon Musk admits that annotation is a tedious job, and it requires skills and training, especially when it comes to 4D (3D plus time series).
Common applications include Point annotation, line annotation, boxing annotation, 3D Point cloud annotation, scene semantic segmentation, and POI (Point of Interest) annotation.
ByteBridge is a human-powered data labeling tooling platform. It provides scalable, high-quality training data for ML/AI industry with flexible workflow.
- The real-time QA and QC are integrated into the labeling workflow as the consensus mechanism is introduced to ensure accuracy.
- Consensus — Assign the same task to several workers, and the correct answer is the one that comes back from the majority output.
- All work results are completely screened and inspected by the machine and human workforce.
In this way, ByteBridge can affirm the data acceptance and accuracy rate is over 98%.
A collaboration of the human-work force and AI algorithms ensure a 50% lower price compared to the conventional market.
Self-driving technology is going to transform the transportation industry, social and daily lives. It’s hard to know when that day will arrive. As life is priceless, we have to seek perfection from the beginning.
As the quality of the labeled dataset determines the success of the self-driving industry, cooperation with a reliable partner can help developers to overcome the data labeling challenges.
We can provide personalized annotation tools and services according to customer requirements.
If you need data labeling and collection services, please have a look at bytebridge.io, the clear pricing is available.
Please feel free to contact us: firstname.lastname@example.org
1 How Auto-Driving Achieved through Machine Learning?
2 Labeling Service Case Study — Video Annotation — License Plate Recognition
3 High-Quality Training Data for Autonomous Cars in 2021
4 What is Semantic Segmentation, Instance Segmentation, Panoramic segmentation?
5 How Data Labeling and Annotation Services Empower Self-Driving Bus?
6 How Data Labeling Service Empower Auto-Driving Industry2021?