Introduction: "Smart car" is an industrial product, before going on the road, must go through sufficient testing, and pass third-party verification, and have a clear description of performance characteristics. At present and in the next few years, no matter what kind of "smart car", the key to judging its "smart" degree lies in whether it improves safety. Guo Min, a traffic engineer, discussed this topic and further introduced how "smart cars" can improve safety.
What are the types of "smart cars" of the future?
In the future, "smart cars" mainly include automatic driving, advanced assisted driving (ADAS) and connected vehicles. Vehicles sold in the future may or may have one of these functions. In the automotive industry, the development process of these three types of functions is not the same as the current situation. Here's a quick explanation.
The first category: self-driving cars. The Society of Automotive Engineers (SAE) defines 6 levels of driving automation (see Figure 1), ranging from 0 (fully manual) to 5 (fully autonomous). This method has now become a classification criterion for most countries. The 6 levels of autonomous driving, L4 and L5 are fully automatic without human driver intervention, the difference is mainly in the regional range, L4 is a limited area, such as a city or region can achieve full automatic driving, while L5 has no regional restrictions.
Figure 1: The 6 Levels of Vehicle Autonomy Explained
Category II: Vehicles equipped with Advanced Assistance Systems (ADAS). L2 and L3 vehicles with partial autonomous driving functions are different from L4 vehicles. Low-level autonomous driving capabilities have long been available and are gradually commercially available. This is due to the continued advancement of advanced assistance systems (ADAS) in the automotive industry and has merged with the development of autonomous driving. Many of ADAS's features, such as adaptive cruise, lane keeping, and automatic parking, have been driverless in specific scenarios and have long been a selling point for commercial vehicles.
The current development of autonomous vehicles has two different routes: one is to directly cut into the L4 fully autonomous driving stage, and mature into a direct development model of fully autonomous driving in the L4 stage; the other is the gradual development mode from L1 to L2, and then to L3 and L4. Due to the traffic safety benefits brought by L2 and L3 levels, it is believed that many advanced assisted driving technologies with safety functions will be used in commercial vehicles within a decade. Therefore, it is expected that there will be ordinary vehicles of a certain scale on the road in the future with L2 and L3 level automatic driving functions. The needs of these vehicles must be considered for road intelligence.
The third category: Connected Vehicles, as another branch of new vehicle technology. After more than ten years of development, many of its applications have been approved for sale and use in many countries. In the European Union, in particular, a number of pieces of legislation have been put in place on data privacy, communications, vehicle installation and sale of connected vehicles, with a focus on security and privacy. In view of the dominant position of Germany and France in the automotive industry in the European Union, in the future, it is a trend for networked vehicles to gradually enter ordinary people's homes. According to some optimistic estimates, due to the low cost of increasing the network connection function, the network connection will gradually become the mainstream configuration of newly sold vehicles, and more and more appear on the road. Perhaps within a decade, more than 20% of the vehicles on the road will be connected vehicles.
What are the criteria for judging how "smart" a car is?
At present and in the next few years, no matter what kind of "smart car", the key to judging its "smart" degree lies in whether it is safer, and the specific evaluation criteria are: whether it can reduce the overall number of road accidents, or reduce the severity of accidents, and reduce the death and injury of people on the road. The method is also simple: in a real-world driving environment, the actual use of the new technology is measured. Real-world statistical results determine whether a technology should be promoted or adopted on as much scale as possible.
Some statistics show that if all vehicles use the three basic advanced assisted driving functions (ADAS) functions of forward collision warning, blind spot detection and lane departure warning, highway accidents will be reduced by 16.3%, accidents on other roads will be reduced by 11.6%, and accidents on other roads will be reduced by 45.4% if more complex ADAS systems such as lane keeping assist, automatic emergency braking, night vision functions, etc. are used, accidents on highways will be reduced by 45.4%, and accidents on other roads will be reduced by 27.5% (see Figure 2).
Figure 2: The future of motor insurance - How car connectivity and ADAS are impacting the market, written by A joint whitepaper by HERE and Swiss Re
Based on such research results, the European Union has released a list of 11 new safety features, which it hopes will be sold by 2021, new cars must have these functions, including a number of ADAS, such as automatic emergency braking, lane keeping assist, intelligent speed assist system, reversing camera rear detection system, networking function, etc. The evaluation criterion of "smart car" is safety, and the actual evaluation method is the numerical value of benefits, including financial benefits and social benefits.
However, the Association of Road Safety Insurance Companies (IIHS) expressed concern in a report:
"One of the safety factors to consider is how many driving tasks can be safely handed over to the car to be done automatically without the driver having to solve it?"
"Vehicle designers are struggling with the functional design of ADAS, and if the ADAS function is limited to ensure that the driver continues to participate in driving, the driver may be spat at because the system function is too low; if the ADAS system looks too powerful, then the driver may think that it is autonomous driving and does not improve attention, resulting in risk."
The safety benefits of new technologies will accelerate their installation on vehicles, and more and more countries will issue mandatory requirements for vehicles to be pre-installed with new technologies that are statistically verified safe and effective. However, no matter how "smart" these new technologies may seem, they cannot replace drivers' driving duties. The illusion of autonomous driving capabilities brought about by the current "smart car" has caused many casualties around the world. Whether it's advanced driver assistance systems (ADAS) or connected vehicles, "smart" is only compared to previous vehicles, not to drivers. The current "smart" still has to be put in quotation marks, and the real vehicle is still a machine that needs to be controlled by people. In terms of safety, the current "smart car" is far from smart enough.
How can "smart cars" improve safety?
"Smart cars" are industrial products that must meet the characteristics of safety, reliability, economy, standardization and so on. That is to say, before the "smart car" hits the road, it must go through sufficient testing, pass third-party verification, and have a clear description of performance characteristics.
This is a long process that complex industrial products must go through. The "smart cars" on the road in the next few years may be going through or have gone through such a process. From the perspective of time history, the main characteristics of the car in the next few years can be described. These "smart cars" deal with all the problems that are currently expected; all the actions taken are also the solutions that people have envisioned or prepared. Technology may need to be imagined, but industrial products must be the expected result of a long-term effort in time.
1
How can advanced driver assistance systems improve safety?
For vehicles equipped with L2 advanced driver assistance systems, in specific isolated scenarios, vehicles can work autonomously; L3 can work autonomously through some ADAS combinations to form continuous scenes under limited conditions. ADAS will evolve to L4 level fully autonomous driving in the future, a trend that has also been echoed by many industry experts. However, the current mainstream ADAS has just reached the initial capabilities of L3. It has roughly the following functions:
Forward Collision Warning (FCW) and Automatic Emergency Braking (AEB) to prevent rear-end accidents or mitigate the severity of rear-end collisions. These types of accidents are more common on busy urban roads, or when the speed is unstable on high speeds, and can identify vehicles, cyclists, pedestrians and animals. When a vehicle thinks a collision is likely, the FCW system alerts it through a series of increasingly urgent audible, visual, or tactile alerts. If the driver does not take any action, it is possible to make an emergency but brief braking autonomously, and if no action is still taken, the system will intervene and use the brakes completely, or avoid the collision altogether, or at least mitigate the severity of the collision. In practical use, the combined effect of FCW and AEB requires a shorter parking line of sight than for human drivers.
Active lane change assist system capable of completing complete lane change operations. The driver activates the active lane change system, which identifies and determines whether lane changes are safe, and when there is no vehicle impact and no vehicle approaching from behind, the vehicle takes full control of the steering and lane changes throughout the maneuvering process. The system discerns the meaning of road markings, signs, and affects the timing and process of lane changes.
Lane Departure Warning (LDW) and Lane Keeping Assist (LKA) systems are currently the focus of many manufacturers' publicity, and together with forward collision warning and automatic emergency braking, they are considered L3 level autonomous driving. These two functions were first used in trucks, keeping commercial vehicles in the middle of the lane for a long time, and their use requires control measures to identify lanes and road sections. Therefore, it is closely related to signs and markings.
There are also parking assistance systems, cross traffic warning systems, steering assistance systems, etc., which maintain safe driving by identifying scenes, signs, markings, vehicles, cyclists, and pedestrians.
2
How can connected vehicles improve safety?
Connected Vehicles are vehicles that are connected to services, out-of-vehicle devices, such as other vehicles and road facilities such as traffic lights or rescue centers, offices, or homes. The word networked vehicle has appeared frequently in recent years because it has many possibilities: for example, it can be improved with the city and road network to improve the quality of residents' work and life; improve the efficiency of municipal services through the management of police cars, garbage trucks, buses, etc.; or through the network to understand the construction work area in front of them, congestion points, take countermeasures in advance to improve safety; or through the network of vehicles, combined with medical, repair, transfer and other services, form a seamless connection to solve individual needs. However, due to the complexity of the governance mechanisms of cities and road networks, such explorations are only carried out in a limited number of advanced areas. Connected cars have attracted much attention, not only because of the benefits of driving safety, but because of their connection with various needs and services. Unlike advanced driver assistance systems, the intelligence of connected vehicles is determined by the intelligence of the mechanisms of the city and the road network. That is, cleverness does not lie in machines, but in people and mechanisms.
However, connected vehicles still play a role in improving driving safety, mainly in two aspects:
Connected vehicles have the ability to obtain timely hazard warnings on the road, while providing controlled driving speed, distance, etc. in hazardous conditions, and automatically handling hazardous scenarios with ADAS.
The vehicle's own information, such as location, vehicle conditions, surrounding scenes, tire tire pressure, etc., can be transmitted to the owner or other professional service agencies in time to understand and deal with unsafe problems in advance.
The wisdom of the road is based on the needs of passers-by, including "smart cars". If the "smart car" only talks about how "smart" on paper, but can't say the specific performance and indicator values, it is at most a delicate model, and it is far from the factory mass production link. None of this should be taken into account in the current wisdom of the road. What's more, vehicles or roads, whether clever or intelligent, are just using the truth that people understand, adopting the mechanism designed by people, and handing over the work of transportation to machines. Industry and engineering only believe in the logic that one plus one equals two. We should follow the laws of industrial development and create a "smart car" and "smart road" that has been clearly thought out and engineers have verified the details.
(This article is reproduced with permission from the surging city hall, and the content has been deleted after communicating with the author.) Author:Traffic Engineer Guo Min)
Editor 丨Li Yunyue Liu Lin