About ADAS (Advanced Driving Support System) and AD (Autonomous Driving)

Learn about Autonomous Driving and Logistics
About ADAS
About ADAS (Advanced Driving Support System) and AD (Autonomous Driving)

About ADAS (Advanced Driving Support System) and AD (Autonomous Driving)

ADAS (Advanced Driving Support System) is a part of a big concept of AD (Autonomous Driving).

Autonomous Driving is classified into five levels according to the degree of automation,
Levels 1 and 2 are positioned as "driving support", and so-called " Autonomous Driving " which the system judges is Level 3 or higher.

ADAS is a "driving support function" to the last, and when a human drives a car.
It is a function to support how to avoid accidents and make it comfortable to reach the destination.
The main body is a human who is a driver, and it can be said that it is a function to support it.
In contrast Autonomous Driving above the goal is no longer car, where AI (artificial intelligence) is determined to be mounted without involvement of human.
It is an idea that it arrives at the destination.

■ This page will explain the differences and common points between ADAS and AD from the level of Autonomous Driving and technical point of view.
In addition, we will explain ZMP's approach to ADAS and AD.

New solution service that utilizes Autonomous Driving

1. About the level of Autonomous Driving

Autonomous Driving classifies the degree of automation with the concept of level.

The classification method is based on the definition of SAE (Society of Automotive Engineers) announced by NHTSA of the US Department of Transportation in 2016.

Level 0: No operation automation

All driver tasks (DDT: Dynamic Driving Task) are carried out by the driver.

It does not contain any automated elements at all. (Including cases supported by preventive safety system)

Level 1: Driver assistance

· Limited implementation of the sustainable operational design domain (ODD) by the operation automation system.

· The operation automation system carries out subtasks of operation task (DDT) related to vehicle control in either the front or rear direction and the left / right direction. (Not both)

· The driver expects to implement the rest of the driving task (DDT).

Level 2: Partial operation automation

· Sustainable operation design domain (ODD) limited implementation by operation automation system.

· The operation automation system carries out subtasks of operation tasks (DDT) related to vehicle control in both the front and rear direction and the left and right direction.

· The driver completes the subtask of object and event detection and response (OEDR)

We expect to monitor the rest of the driving task (DDT).

Level 3: Conditional operation automation

· Sustainable and operational design domain (ODD) limited implementation of all operation tasks (DDT) by Autonomous Driving System (ADS).

· Preliminary response users (DDT fallback-ready user) are requested to intervene by Autonomous Driving System (ADS)

We expect to respond appropriately to system failures related to operation task (DDT) implementation in other vehicle systems.

Level 4: Advanced operation automation

· Sustainable and operational design domain (ODD) limited implementation of all operation tasks (DDT) by Autonomous Driving System (ADS).

· In fallback (DDT fallback), users are not expected to respond to intervention.

Level 5: Full operation automation

· Sustainable and unconditional for all operation tasks (DDT) by Autonomous Driving System (ADS)

(Ie, operation design domain (ODD) not limited).

· In fallback (DDT fallback), users are not expected to respond to intervention.

(Source: Ministry of Land, Infrastructure and Transport http://www.kantei.go.jp/jp/singi/it2/senmon_bunka/detakatsuyokiban/dorokotsu_dai1/siryou3.pdf）

2. Positioning ADAS in Autonomous Driving

ADAS is generally positioned at Level 2 of Autonomous Driving. This means that the driver is the decision-maker. In other words, the driver, that is, human is responsible for the accident.

Recently, sales of vehicles aimed at achieving Level 3 have also started, and Honda plans to sell model Legend, which equipped with an Autonomous Driving function on highways.

In Level 3 Autonomous driving, the car is responsible for the accident while driving, and the driver is relieved of driving. At Level 3, when the vehicle cannot continue autonomous driving, the driver has to take over driving promptly, and is expected to play an active role in various areas, such as following a car during congestion on a highway.

(Reference: Jiji News
Japan's first "Level 3" automation by Honda during high-speed traffic congestion in the next summer.https://www.jiji.com/jc/article?k=2019121600906&g=eco）

3. Common points between ADAS and Autonomous Driving

ADAS (Advanced Driver Assistance System) and AD ( Autonomous Driving ) are useful for extracting useful information from various sensors such as in-vehicle cameras and radars that sense information on the outside world, and data measured by sensors.There are many common technologies such as ECU, steering, accelerator, brake, etc, which controls the drive part of the car. ADAS development leads to the development of AD as it is.

The image at the time of advancing ADAS development is to take measures to "eliminate the cause of accident" or "minimize or minimize the damage". On the other hand, since Autonomous Driving is assumed to be unmanned and the system will operate, it is necessary to clear the safety standards beyond human driving. The fact that equipped with highly functional control equipment capable of Autonomous Driving makes ADAS functioning when driving a human being a very high standard.

4. Difference between ADAS and Autonomous Driving

The fundamental difference between ADAS and Autonomous Driving is the location of responsibility for driving.

Among Levels 1 to 5 of Autonomous Driving , there is a responsibility to humans up to Level 2, which is ADAS.
At Level 3, if the operation can not be continued, the driver responds and becomes the main entity, but basically the system has the subject, it becomes Autonomous Driving .

In both Level 2 and Level 3, the system may perform the driving operation of the car such as accelerator, brake, steering, etc.
However, in Level 2, the driving is done only by the driver, and in Level 3, depending on the human support conditions, when the system is controlling the vehicle, the main driver is the system.

And at Level 4 and above, the driving is completely system driven.

When humans are responsible, it is the same as in the past, but when an accident occurs under the Level 3-5 system, it is not clear at present who will take responsibility for the system.

In Japan, the responsibility of a car accident should be considered from a criminal, civil, or administrative perspective, but the current systems and laws are based on the assumption that humans will drive, and there will be no driver.

In the future, it is necessary to develop a legal framework in order to make full use of Autonomous Driving in full scale and to spread it.

For example, consider paying insurance for accidents that occurred during Autonomous Driving .
If the accident happens due to the car maker's negligence, it is expected that the insurance company will make claims against that manufacturer.
In this way, discussion on practical application of Autonomous Driving and legal improvement is active actively in each direction.

5. HMI (Human Machine Interface) of ADAS

Another difference between ADAS and AD (Autonomous Driving) is that while Autonomous Driving operates only by the system, in ADAS both humans and systems are involved in the operation. This is where the two need to be bridged.

Since humans and the system operate the car jointly, it is necessary for both parties to communicate smoothly. The point to keep in mind when designing ADAS is the idea of the human machine interface (HMI). It can be said that this is a protocol in which humans and machines communicate their intentions and information.

When a person sends a command to a car, the car detects the driving operation itself of the person.
On the contrary, the transmission from the car to the human being is visual display on the monitor, auditory sense that sounds a warning sound, tactile sense such as steering vibration.
In order for human to accurately recognize the information transmitted from the system, the design of the interface becomes important.

Information transmitted from the ADAS system to the driver is classified into three levels (information provision, warning, alarm).
In the early stage, when the system communicates information that the driver can not directly confirm,
If you do not tell the intent accurately, the driver misunderstands and it is dangerous.

If you do not decide the audiovisual display in an appropriate form, there is a risk that the operation will be delayed by mistake.
For example, in a situation where there is no time delay, there is a limited time to view, so it is necessary to notify the information immediately by a signal sound and instantaneously cause the driving behavior. In this way, we will endeavor to adopt a transmission format appropriately transmitted according to the situation.

Also, drivers with low proficiency in driving and without sufficient prior knowledge on ADAS functions are required to design interfaces that can move to appropriate actions. When designing the Human Machine Interface (HMI), it is necessary to consider the age of the driver and proficiency in driving, especially for older people and beginners of driving.

6. ADAS and Autonomous Driving in intelligent transportation systems

Although Intelligent Transport Systems (ITS) that attempts to construct human, vehicle, and transportation infrastructure as comfortable, safe and highly efficient transportation systems using information and communication technologies, Autonomous Driving and ADAS.

The birth and mass production of automobiles has brought about a great revolution in migration, but since then it has predicted that the fundamental mechanism of human driving that has not changed for centuries , will change dramatically. Along with recent remarkable evolution of IT, IOT penetration, networking, AI (artificial intelligence) deep learning (deep learning), algorithms etc., systemization of Autonomous Driving is progressing rapidly.
In other words, the driving of a car will be switched from human to ADAS, and to the Autonomous Driving that ADAS has developed.

Changes in the dangerousness of driving of a car due to driver's condition change that can not be avoided by living human beings such as good or bad physical condition, aging, misunderstanding and concentration of concentration will be stabilized by the evolution of ADAS and Autonomous Driving. As ADAS and Autonomous Driving technology become more sophisticated, car movement will be more comfortable, the risk of accidents will be reduced, and road traffic will be safer.

Moreover, it is expected that not only traffic accident reduction, but also efficient traveling will reduce congestion and environmental burden, and will contribute greatly in solving problems in road traffic society. Autonomous Driving enters the practical stage, people who were previously devoted time and effort to driving are relieved from work, leisure is born, and dramatic changes in industrial structure are brought about.

Advancement has advanced further with the data base such as traffic data generated according to the Autonomous Driving and ADAS technology, the Autonomous Driving System becomes data driven type, and as the fusion of the Internet with the Internet progresses as symbolized by a connected car, AI (artificial intelligence) and software technology weights are on the rise.
As a platform indispensable for its progress, utilization of cloud services will also be emphasized in order to process huge amounts of data such as running video database, dynamic map etc.

Each car maker, such as Volkswagen, Toyota, Renault, Nissan, Mitsubishi, General Motors, Honda, Peugeot Citroën, Daimler, Mazda, Subaru, Daihatsu, Suzuki etc. make many investments, we are concentrating on how to raise added value to consumers and maintain and strengthen competitiveness by taking into account the trends of new technologies represented by ADAS and Autonomous Driving on vehicle manufacturing technology.

7. ZMP's involvement with ADAS and AD

ZMP is working not only on automobiles, but also on automating various "moving objects".
We are also promoting automation of not only general cars but also taxis, logistics robots, home delivery robots, drone, etc. where autonomous movement is required.
ADAS (Advanced Driving Support System) and AD (Autonomous Driving) take three steps of recognition, judgment and operation.
We develop our own "brain" that is responsible for judging "eyes" responsible for cognition among them.

For both ADAS and AD, an important eye stereo camera for acquiring external information is a high sensitivity CMOS imager.
From pre-processing such as synchronization / parallelization / distortion correction, distance detection from parallax, object detection, etc., we are developing our own company.
For example, even with the same ADAS function, recognition and prediction such as other traffic and pedestrians on urban driving and expressway are greatly different.

In ZMP, practical running test of public roads in urban areas is practiced, accumulating huge amount of data and experience.
We are also working on technology development aiming at Autonomous Driving Level 4.
As a matter of course, ADAS compliance is advancing the development of new devices and algorithms so as to realize AD and more advanced Autonomous Driving .

8. Introduction of ZMP Autonomous Driving Development Solution

In ZMP, we are aiming for Level 4 autonomous driving while drawing a "society where automated driving is put to practical use in the future". We are working on technology and development of taxi dispatch applications.

In addition, we are engaged in business development through service development,through our own autonomous driving control development and applying technology developed through technological development to products, which enables us to use it as a development tool for ADAS / autonomous operation.

Below, we introduce the service utilizing Autonomous Driving technology.