A brief discussion on the future development and challenges of autonomous parking systems

Talking about the current autonomous driving technology, the most familiar thing to the public must be the assisted driving function that has entered the homes of ordinary people. Compared with L4 or higher level autonomous driving, drivers believe that the steering wheel can be in their hands at all times and they can take over in time if there is an emergency. Among the assisted driving functions that consumers can currently experience, autonomous parking is definitely one of the most technically difficult and practical functions. Imagine that a novice driver who has just obtained his driver's license needs another driver to help guide him while reversing the car into the garage in a narrow parking lot. At the same time, he must be careful to avoid being scratched. The implementation of autonomous parking technology will solve this problem that has troubled novice drivers for many years, making parking no longer painful and more convenient. This article will lead readers to understand the most popular assisted driving function at present-autonomous parking.

What is autonomous parking?

The autonomous parking system is referred to as APS, which is the abbreviation of English Automated Parking System. Cars equipped with automatic parking functions do not require manual intervention. With the help of on-board sensors , processors and control systems , they can automatically identify parking spaces and automatically complete the parking process.

When using the autonomous parking function, the driver can get off the car at the designated drop-off point and issue parking instructions through the mobile APP. After receiving the instructions, the vehicle can automatically drive to the parking space in the parking lot without user control and monitoring. You can also give a car pick-up instruction through the mobile APP. After receiving the instruction, the vehicle can automatically drive from the parking space to the designated pick-up point. While the vehicle is driving autonomously, it can still abide by road traffic rules or on-site traffic rules set by the parking lot operator. Generally speaking, autonomous parking mainly includes three parts: finding a seat, parking and calling a car.

The development history of autonomous parking

The automatic parking system can be traced back to 1992, when Volkswagen used automatic parking technology on its IRVW (Integrated Research Volkswagen) Futura concept car . IRVW is a car with fully automatic parking function. The driver can get out of the car and watch the whole process of automatic parking. A PC -sized computer is installed in the trunk to control the entire automated parking system. Volkswagen estimated at the time that this feature would increase the price of the car by about $3,000, so it did not put the system into production.

The automatic parking function became truly commercialized when Toyota provided an optional automatic parking function on the Japanese Prius hybrid car in 2003. Three years later, British drivers could add it to their Prius for $700. Automatic parking function.

In 2004, a group of students at Sweden's Linkopeng University collaborated with Volvo on a project called Evolve. Evolve cars can automatically park in parallel. The students installed sensors in a Volvo S60 and a computer in the trunk that controls the steering wheel and accelerator and brake pedals. Siemens VDO is developing an independent driver assistance system called ParkMate that can help drivers find a parking space and park their car.

Nowadays, many high-end models are equipped with autonomous parking functions, most of which are optional systems. The most well-known among domestically produced cars is Xpeng's autonomous parking system. It was well received by consumers once it was launched and will be one of the key technologies leading the way in intelligent cars in the future.

Autonomous parking technical details

The automatic parking system consists of multiple sets of sensors, such as radars, cameras, etc. all over the car body. After collecting relevant data such as images, distances, etc., the data will be transmitted to the processor. The processor will analyze and process the collected data to form an automatic parking system. The parking strategy then converts it into a signal . After receiving the signal, the control system controls the car to perform actions such as steering, reversing and braking according to the instructions until the parking process is completed.

The automatic parking process can generally be divided into three parts: parking space detection, path planning and path tracking.

During the first step of parking space detection, the vehicle needs to maintain a low speed. At this time, the ultrasonic radars on both sides of the car start to work and continuously scan the parking spaces on both sides of the parking space. When the distance between the scanned parking spaces is enough for the vehicle to park, the search ends.

After determining the parking space where you need to park, the second step is to plan the path from the current location to the parking space. This step is the most important and difficult step in autonomous parking. The difficulty of this step is that we first need to know the precise location and orientation of the parking space, and then generate the driving path through the corresponding planning algorithm . During the generation process, we need to consider the obstacle avoidance actions of static obstacles and the dynamic constraints of the vehicle to ensure autonomy. Parking safety and comfort. During the parking process, the vehicle must also respond to dynamic obstacles, predict the future trajectory of obstacles at low speeds and short distances, and plan the route speed in advance to avoid them. This is also an indispensable function in path planning.

After the final trajectory is generated, path tracking is performed. A good tracking algorithm can make feedback adjustments at all times to ensure the safety of the final actual driving path when there is a certain error in positioning and the actuator is not executed in place.

It can be said that autonomous parking technology requires vehicles to have sufficiently intelligent perception and decision-making planning systems to achieve L3 or even L4 level autonomous driving functions. At the same time, the parking lot also needs to have high-precision map coverage or autonomous mapping functions so that vehicles can be parked in place. This also represents the highest level of current mass-produced assisted driving functions.

Future development and challenges of autonomous parking systems

As automobile technology continues to improve, automatic parking systems will make technological breakthroughs and evolve toward autonomous parking. Autonomous parking, also known as valet parking or one-touch parking, allows the driver to summon a vehicle in a parking space at a designated location or park the vehicle currently driving in a designated or random parking space. The entire process does not require human operation and supervision under normal conditions, and corresponds to the L3 level of SAE autonomous driving.

However, at this stage, the autonomous parking system is still not functionally perfect. Due to the high requirements for sensing hardware and decision-making software, there is still a long way to go in terms of technical implementation. At the same time, due to the incomplete coverage of high-precision maps in most parking lots, current parking only remains in the last short distance of reversing, which also greatly limits the function of this function to a certain extent.

In the future, autonomous parking will be used by everyone as a technology with the highest frequency of contact with consumers. For example, by tapping the phone before using the car, the vehicle can arrive at the predetermined location in advance. After arriving at the destination, the vehicle can be parked by tapping the phone. The vehicle automatically parks into the parking space. Autonomous parking will be widely used by everyone before autonomous driving becomes popular, truly solving the problem of drivers having difficulty parking.