Choose an AI Instructor When Taking a Driving Test, He Will Be Calm and Won’t Curse

The coach is the nightmare of countless driving school students. If you are unlucky enough to meet a coach with a bad temper and who likes to curse, it is really painful. Fortunately, many driving schools across the country have gradually launched AI coaches.

Recently, a video of a driving school in Shanghai using an AI coach to teach became a hot search on Weibo.So far, there are more than 2,100 students.With the help of the school's AI coach, he successfully graduated.

The driving school adopts a combination of software and hardware to improve the traditional driving school practice model.

In terms of hardware, the training vehicle is equipped with a GPS locator and radar.Two high-precision GPS locators can monitor the position of the training vehicle in real time, and multiple radars installed at the front of the vehicle are used to detect obstacles.

In terms of software, the AI coach will also create a personal profile for each student.Assess students' training ability and adjust exercise content based on their learning situation.

In addition, the car is also equipped with an active braking function to ensure the safety of students when practicing alone in unexpected situations.

However, behind the driving schools' constant attempts at new technologies and breakthroughs, there is also a sense of anxiety that they have no choice but to do so.

Driving schools are competing fiercely, and cost control is a problem

With the development of social economy and the improvement of people's living standards, the number of private cars has shown explosive growth since the reform and opening up.At the same time, driving schools are springing up all over the country.

According to data released by the Ministry of Transport,From 2017 to 2018, the number of driving schools in China increased by 7.1% year-on-year, but the number of students showed a downward trend, down 0.8% year-on-year.

The imbalance between supply and demand in the driving training market has further intensified competition in the industry.Registration price, training venue, condition of the training vehicle, waiting time for practice...These have become important factors in driving schools competing with each other and snatching students.

However, due to high operating costs, some driving schools are unable to improve user experience and hardware equipment.

The Qualification Requirements for Motor Vehicle Driver Training Institutions (GB/T30340) stipulates that:The site area of a third-level driving school shall not be less than 10,000 square meters.For each additional small vehicle, the site area should be increased by 400 square meters.

In 2019, the Ministry of Transport issued the "Regulations on Motor Vehicle Driver Training Management (Draft for Comments)", which clearly stipulates the number of training vehicles for driving training institutions.Taking the third-level driving school as an example, it is required to have no less than 20 teaching vehicles.

In addition, when an ordinary driving school hired a coach,The same number of coaches as training vehicles is required, and the coaches also need to participate in various trainings regularly.In addition to investing time, you also need to bear certain training costs.

The high site rental fees, high training vehicle fuel and maintenance costs, plus labor costs,These have become the "three big mountains" weighing on driving schools.

Technology: A lifeline for driving school upgrades

Faced with high operating costs,Some driving schools choose to replace fuel-powered training cars with new energy electric vehicles.

According to Wang Xiangyang, the person in charge of Jiaduoduo Academy, the use of new energy training vehicles will reduce the fuel costs of 70%-80% and the maintenance costs of 50%. However, the replacement of training vehicles with new ones is also an expense that cannot be ignored.

Some driving schools are pinning their hopes on technological upgrades.

•  AI Simulator 

The AI simulator is a device for practicing road tests. It looks like a racing game console.Equipped with a curved display and steering wheel, it can simulate a real driving experience.

The road test environment simulated on the curved screen is very realistic, including roads, traffic lights, crosswalks, trees and buildings, and can even simulate driving conditions on different routes and in different weather conditions.

•  AI Driving School Examiner 

Microsoft Research launched a safe driving project called HAMS in 2019.

HAMS stands for Harnessing AutoMobiles for Safety.Early HAMS was designed to monitor the driver's status with the help of AI technology and multiple sensors to ensure driving safety.

Then,HAMS is being used to invigilate driving tests in a location in India. ALT stands for Automated Driver License Testing, which refers to fully automated driver license test proctoring.Through the front camera of the smartphone, determine the identity of the reference person, whether the seat belt is fastened, whether the rearview mirror is checked when turning or overtaking, etc.

The rear camera is used to track the vehicle's trajectory to ensure that the vehicle does not deviate from the designated track during test items such as parallel parking and continuous curves.

AI is not omnipotent and still has many limitations

The involvement of artificial intelligence technology has accelerated the upgrading of traditional driving schools. Taking the AI coach introduced in the news at the beginning of the article as an example, the labor cost of driving schools will be greatly reduced.

Data released by driving training robot manufacturer Yixian Technology shows, one human coach can manage 30 AI coaches, saving the driving school 95% in labor costs every year.

In addition, replacing human coaches with AI can also improve user experience.Previously, many videos of students being scolded by driving instructors had been circulating on social media, and many netizens also left comments under the videos of AI coaches, saying that "it is friendly to people with social phobias", "you no longer have to worry about being scolded by the instructor", and "the psychological pressure is reduced".

However, in the early stages of the road test, such as the second and third subjects, the limitations of AI still require more involvement of human instructors.

 1. In the introductory stage, scientific guidance from human coaches is still needed 

For novice trainees who are not familiar with motor vehicle operation, human instructors are more suitable to guide them.Accurately and comprehensively inform the use and precautions of motor vehicles.

In comparison, AI coaches are more suitable to join in the middle and late stages of driving practice. On the premise that students have mastered a certain foundation, they can enhance and improve operational skills such as reversing into a parking space and starting on a slope.

 2. Students are overly dependent on screens and become distracted 

AI coaches often install a display screen in the car to update the vehicle's driving status and whether it crosses the line in real time. This helps students understand their own practice situation and find out what they are missing.

But long-term excessive focus on screens may have the opposite effect.As a result, I was not used to looking at the rearview mirror during practical exercises, which led to my failure in the exam.

 3. The coach cannot impart theoretical knowledge during the practical operation 

In road test practice, human instructors can often provide targeted teaching and actively teach theoretical knowledge during the course.Interspersed with practical driving skills beyond the scope of the test,Such as defensive driving methods and emergency response methods for dangerous situations on the road.

This knowledge may not be tested in the exam, but it is very practical, and AI coaches are currently unable to do this.

Every driving school student is a participant in future traffic safety.The goal of a driving school should not just be to help students pass the exam and get a driver’s license.

The involvement of AI technology can improve the efficiency and effectiveness of driver training to a certain extent, and help students learn and practice driving more scientifically.However, without the involvement of human coaches and the teaching of practical skills,I'm afraid that some "experienced drivers" may have to search on Baidu on how to open the hood even though they have been driving for two or three years.