The 5th Meeting of the Sub-Working Group on Social Rules for automated driving vehicles in age of AI

Overview

Date and time: April 26, 2024 (Friday) from 10:00 to 12:00
Location: Digital Agency Online
Agenda:
1. Opening
2. Proceedings
  1. The administrative legal Issue of automated driving
  2. Opinions, etc. of the Association of Bereaved Families of Seki Higashi Kotsu Crimes (Ai No Kai)
  3. Explanation by the Secretariat (regarding the main opinions, the direction of the compilation, and the future schedule of the 4th age of AI Sub-Working Group on the Review of Social Rules for automated driving vehicles)
3. Exchange of opinions
4. Adjournment

Materials

Minutes

Councilor Kodama: First of all, I would like to make an administrative contact. Today's meeting will be held in a hybrid format, face-to-face and online. All members who will participate in person will be asked to raise their hands before making their statements. All members who will participate online will be asked to turn on their cameras during the meeting and unmute their microphones when they speak. In addition, we would like to ask you to mute the cameras and microphones when other people are speaking. We would also like to ask the audience to turn off their cameras and microphones.

Next, I would like to check the materials. As stated in the agenda sent in advance, the materials include the agenda, the list of members, the explanatory material for Member Harada, two types of explanatory materials for the Seki Higashi Kotsu Crime Bereaved Families Association (Ai no Kai), the explanatory material for Member Takahashi, the explanatory material for Mr. Kuki Honda Motor Co., Ltd., the explanatory material for the Secretariat, the materials to be submitted by the members, and the list of attendees. If you do not have them, please contact the Teams chat function or the Secretariat by e-mail.

As for the attendees today, due to time constraints, I would like to distribute the list of attendees in your possession instead of introducing them. In addition, I heard that Ochiai members are scheduled to participate in the middle of the meeting. This time, Mr. Matsunaga, Vice President of the Association of Bereaved Families of Seki Higashi Kotsu Crime (Ai no Kai), and Mr. Kuki, Chief Engineer of Honda Motor Co., Ltd., are attending as guest speakers. In addition, Mr. Ohno and Mr. Suganuma from DADC are also attending as guests, as in the previous meeting. Please be aware that the materials and minutes of the plenary meeting will be made public at a later date.

Then, I would like to ask Mr. Kozuka to proceed from here. Mr. Kozuka, please.

Mgr. Kozuka: Thank you very much, . Thank you very much for meeting face-to-face and online today. I will proceed according to the agenda.

First, I am Administrative and Legal Issue of Item 2 - (1) automated driving. Then, Mr. Harada, I would like to thank you for your cooperation.

Member:
(The following explanation is based on "Material 3: Administrative and Legal Issue of automated driving")
Thank you very much. Today, I would like to explain about the administrative legal Issue of automated driving. What you see now is the external view of the problem. There are many administrative legal Issue surrounding automated driving. Today, I would like to focus on these three. The first is the responsibility of the driver, which is related to the Road Traffic Law. The second is the responsibility of the manufacturer, which is related to the Road Transport Vehicle Law. The third is the legal system related to roads as public domain or infrastructure. It includes Article 2 of the State Redress Act on public domain management and liability for damage in the event of accidents. However, this time, it is not a matter of State Redress Act, but the distribution of responsibility in the event of accidents in automated driving. Please turn to the next page.

With regard to the liability of drivers, as is well known, the current Road Traffic Act is based on vehicles and focuses on the obligations of drivers or users. In particular, the majority of the provisions set forth the obligations of drivers, and the driving license system is at the center of this. In addition, compared to other administrative law systems, it is characterized by an extremely large number of sanctions based on criminal penalties. And by combining this with the penalty system, it has extremely high effectiveness as a means of ensuring the performance of obligations. On the other hand, if a complete automated driving is realized, the concept of the driver naturally disappears, so there will be no one to whom the obligations related to this are addressed, and the premise of non-criminal processing centered on criminal penalties will be lost. Therefore, one option is to reorganize the obligations of actions by focusing on the user rather than the driver, and to achieve a certain effect by linking it to, for example, the refusal of automobile inspection, like the current abandonment penalty. It is written that the abandonment penalty is different from the penalty, and if the obligor does not have the ability to pay, the final performance of obligations cannot be secured. The other is to reorganize the obligations of actions by the manufacturer of the automobile and change it to criminal penalties for the manufacturer. However, compared to the driver or user of the automobile, the manufacturer manufactures many vehicles, so it will be extremely wide, and if the criminal penalties are combined, I think there is a possibility that a debility effect will occur.

So, next is the responsibility of the manufacturer. The Road Transport Vehicle Law stipulates safety standards for the structure and equipment of automobiles, and users are not allowed to operate vehicles that do not meet these standards. In addition, it is possible to recommend that automobile manufacturers take improvement measures in order to make automobiles that are deemed to be at risk of not conforming to the safety standards conform to the standards, and it is also possible to make further progress through public announcement and orders. In addition, compliance with safety standards is also a requirement for designation in relation to type designation, and if this is not met, the designation will be canceled. Administrative law such as this security standard has a certain meaning in relation to civil liability. In particular, obligations under the administrative law are positioned as one of the elements of the duty of care, which is an extremely important factor in determining the success or failure of compensation for damage based on tort, and it can be said that if there is a violation of obligations under the administrative law, it is likely to be determined that there is a violation of the duty of care. On the other hand, the duty of civil care is not limited to compliance with administrative law, and other factors may be considered, so it is not a relationship in which if you comply with administrative obligations, you will never be liable in civil matters. When a complete automated driving is realized, it is conceivable that the concept of the driver will disappear, and the roles of the manufacturer will be further expanded. In addition, due to the rapid pace of technological innovation in automated driving, it is likely that security standards will need to be updated. Some administrative laws, such as the Building Standards Law, take the method of fixing safety standards at the start of construction of facilities, so-called existing nonconformance, but there are also cases where compliance with revised standards is required, such as the regulation Act on Nuclear Reactors, etc. in after amendment in 2012. Therefore, it is considered that the Road Transport Vehicle Law should be provided with provisions that assume immediate application in the same manner as the regulation Law for Nuclear Reactors. In addition, if the security standards themselves are not promptly updated, there is a possibility that the administrative side will be held accountable for not exercising the authority under Article 1 of the State Redress Act. This position has been expressed in several judgments, including the Chikuho Pneumoconiosis Case of the Supreme Court. Furthermore, given the current situation in which manufacturing in automated driving vehicles is not necessarily limited to domestic manufacturers, I think it will be necessary to discuss extraterritorial application. Even now, the Road Transport Vehicle Act has a provision on designated foreign manufacturers, etc., and of course, in that sense, extraterritorial application is assumed, but it can be fully assumed that there will be difficulties in actually enforcing it due to the jurisdiction of enforcement. Therefore, one of the recent legislative examples that can be used as a reference is a method in which domestic representatives or agents are established and administrative investigation or administrative dispositions are made to them. In addition, it may be considered to institutionalize a system related to the safety of ships that combines voluntary regulation. Please take the next page.

Finally, I will explain the new way of responsibility distribution and behavior control in the world of perfect automated driving from the viewpoint of administrative law. Regarding current automobile accidents, basically the responsibility of the driver is centered, and compulsory automobile liability insurance is linked to it. In addition, if there is a problem of road equipment, it is a problem of construction liability under Article 2 of the State Redress Act, and there is also a problem of liability for non-exercise of authority under Article 1 of the State Redress Act. Or, although it is very exceptional, there is a possibility that criminal liability of public officials will be pursued.
However, in the perfect world of automated driving, it is difficult to identify the cause of the incident or the perpetrator, and the predictability and possibility of avoiding the consequences are judged in a subsequent trial, so it seems that there are difficult problems. Therefore, we focus on the cause of the accident, which is the current method, and determine whether or not the person is responsible as an individual. This is called digital and micro-responsibility allocation here, but if this is done, it will result in the person who happened to be closely related to the accident site taking all the responsibility, or it will be impossible to eliminate the possibility that various chilling effects will be caused by fear of this. Therefore, our basic image is not of all the accidents in automated driving, but of automated driving, which has the characteristics I just mentioned and for which it is difficult or impossible to identify the causative acts and actors, we are considering establishing a so-called compensation law as the current tort law or a special law for State Redress Act, and implementing an analog distribution of liability in which the costs are statistically distributed while maintaining the level of compensation for such accidents to date. First of all, although the user of an automobile should not normally be responsible in the form of a breach of duty, it is possible to pay attention to the fact that he / she receives daily benefits from automated driving vehicles and to assume a certain financial burden as a so-called beneficiary burden. In addition, the automobile manufacturer will pay the contribution in terms of liability and risk. Furthermore, with regard to the possibility of pursuing the responsibility under Article 2 of the State Redress Act in road or road traffic administration and the possibility of pursuing the responsibility under Article 1 of the State Redress Act in automobile safety administration, I have an image of securing financial resources for benefits for victims by making contributions. When an incident occurs in automated driving, the cause of the incident will be analyzed by the organization. However, the results will not be reflected in the cost sharing of benefits for the incident, but will be positioned as a statistical indicator for changing the macro cost sharing ratio in the next fiscal year. In addition, automobile manufacturers are considering a mechanism in which they receive safety evaluations of vehicles and programs from the Safety Evaluation organization and their contributions fluctuate depending on the results. This is intended to provide an incentive for automobile manufacturers to produce safer automated driving vehicles vehicles by incorporating behavioral change using an insurance system, and to clear the so-called barrier of extraterritorial application by making the organization for safety assessment independent from the public authorities of the country. Of course, it costs a lot to create such a large-scale system, and I think the feasibility is low, but I introduced it to the effect that if you draw an ideal diagram ignoring various assumptions, there could be such a system. That's all.

Mgr. Kozuka: Thank you very much, Harada. Next, let's move on to agenda 2 - (2). I heard that the Seki Higashi Kotsu Crime Bereaved Families Association (Ai no Kai) will give us your opinion. I heard that Vice President Matsunaga will give us an explanation. Thank you, Mr. Matsunaga.

Mr. Matsunaga:
(Explanation based on "Material 4: Written Opinion of the Association of Bereaved Families of Seki Higashi Kotsu Crimes (Ai No Kai)")
I am Matsunaga, Deputy Representative Director of the Association of Crime Bereaved Families of the Higashi Kotsu. Thank you very much for inviting me to this review meeting. We are commonly called Ai no Kai, but Ai no Kai is a group of bereaved families who lost their loved families in traffic accidents. I myself am a bereaved family who lost my wife and daughter in an accident in April 2019 in which an elderly driver mistakenly stepped on the accelerator and brake. I have submitted a written opinion at this review meeting, and I will read it out.

First, since the spread of automatic cars, there have been a number of accidents caused by the elderly, including accidents in which the elderly mistakenly step on the pedals, and accidents in which the elderly are unable to make instantaneous decisions or take action due to advanced dementia. In such cases, it is the weak in traffic and good citizens who follow the rules that are always affected.

However, just calling on the elderly to voluntarily return their licenses will not reduce the number of accidents. This is because in "rural areas" there is a great shortage of ways to get around. In urban areas, subways and trains are spread out in a network, and buses run frequently. However, in rural areas there are places where there are no subways or trains, and there are only a few buses running a day. It takes an hour to walk to a convenience store, and it is an environment that is unthinkable in a city. In addition, for people involved in agriculture and forestry, not being able to drive a car is a matter of life and death. automated driving is a "star of hope" that can solve these serious problems.

In addition, it will solve the shortage of labor in the truck and bus industries, taxis, and so on, which will be beneficial to the economy of the entire nation. Above all, traffic accidents will be dramatically reduced. It will save the lives of many of the people who are most important to you. For both the victims of traffic accidents and the entire people, automated driving vehicles has the highest expectations of automated driving. Therefore, I would like to see Tokyo promoted by all means.

Next, I believe that we should consider systems based on the premise that accidents will not completely disappear even in automated driving.

Based on this premise, what measures are available to reduce the number of accidents to as close to zero as possible?

One possible way is to oblige automated driving vehicles to install external airbags.

Second, I think more fundamental measures are necessary. Environmental improvement of roads. For example, I think that it is possible to ensure that pedestrians and vehicles are separated at intersections, to lower the speed limit in residential areas, to create a dedicated lane for automated driving in urban areas in the transition period until most cars are converted to automated driving vehicles, and to make signs and traffic lights more easy-to-understand.

Finally, I would like to say one more important thing. It is a "major premise" story to bring the number of traffic accidents close to zero.

We, who are amateurs in law, may be preaching to you, but I would like to say this because it is very important for the bereaved family of the traffic crime.

Even if it is automated driving, please create a system in which the engineers who make it learn the law properly and follow the traffic rules firmly.

Cars today are driven by people. People inevitably make mistakes. Because they forget the traffic rules they learned at the driving school, or even if they remember them, they don't follow them before they know it. That's why accidents don't disappear.

Then, how about the case of automated driving? automated driving is not driven by people.

So, are the machines driving on their own? Or are they being driven by God? They are both different. A system that performs automated driving is programmed by humans, and "humans are driving the machine." What if, when creating a automated driving program, the person who creates it does not know the traffic rules? As a result, automated driving cars will run on public roads without observing the traffic rules. If this happens, a major accident will occur someday, and new victims will be born. This automated driving is intended to eliminate accidents as much as possible, but on the contrary, it may lead to the fact that accidents do not decrease.

In addition, if such accidents happen and engineers and manufacturers who make vehicles that cannot follow traffic rules are not subject to criminal punishment at all, I don't think the people will be convinced. On top of that, I think no one will trust such a automated driving and they will stop riding it because they are scared. As a result, automated driving may decline rather than spread.

Then, what is that "traffic rule"? It is a "road traffic law" that is familiar to us. Therefore, I would like manufacturers to learn the road traffic law properly, program a system that can firmly follow the rules, and make a automated driving vehicles. If a car is made in such a way, accidents will be as close to zero as possible.

In addition, I have heard from the advisor of the Ai no Kai that if an accident occurs even if you observe the Road Traffic Law, it means that an accident can occur even if a person is driving. Therefore, criminal liability will not be charged.

Of course, we would like automated driving to be promoted, but first of all, we would like the road environments involved in it to be improved. Even in today's society, if all people in society firmly observe the Road Traffic Law, accidents should be as close to zero as possible.

However, as long as there are people who forget or do not obey the Road Traffic Law, accidents will occur. Therefore, I think the same thing can be said in automated driving. If you do not observe the Road Traffic Law and do not stop by a sensor because a person comes in front of you, but if you do programming to firmly observe the Road Traffic Law, accidents will occur less frequently, and if accidents occur, we believe that criminal liability will not be charged. That's all. Thank you very much.

Mgr. Kozuka: Thank you very much, . I learned a lot from your comments. Thank you very much.

Next, in relation to this, Mr. Takahashi, a member, has submitted materials. Mr. Takahashi, could you please explain the materials?

Takahashi Member:
(Explanation based on "Material 6: What is Necessary for automated driving Transformation" below)
This is Takahashi. Before the start of this study meeting, I had only a vague idea, but as I asked various teachers to submit materials and listen to their stories, my mind was gradually being organized, and I learned a lot. When I think about where we end up, I think it's the relationship between predictability and the duty to avoid consequences. Certainly, when we make a car, I think we have been making it centered on such things as how people are here, how they move, and what kind of action they are going to take here. Predictability is limited beyond what humans can program. In what cases accidents have occurred, the Road Traffic Law is the culmination of such things, and it is a judicial example. In the Road Traffic Act, the content of the duty of care, mainly the duty to avoid consequences, is specifically written. The Duty to Avoid Consequences is the duty to do this because this is the situation, because this is foreseen and foreseen. Such obligations are detailed in the Road Traffic Law. In other words, even in the case of automated driving, if you properly follow the obligations written in the Road Traffic Law to do such things, accidents do not occur in most cases. Manufacturers may think from the perspective of how to respond to dynamic phenomena such as the presence of a person in front of them or the occurrence of some event. That's certainly important. However, I would like you to look back once and learn about the Road Traffic Law and court cases, which are the culmination of what Japan has made so far. It corresponds to the case where it can be roughly predicted when this is observed. Therefore, if the programming properly observes the obligation to avoid consequences, in other words, the road traffic law, even if an accident occurs, it is not responsible because it was originally unpredictable. On the other hand, if you create a programming that does not comply with the Road Traffic Act, when an accident occurs, even though it was predictable in the first place, the accident occurs because you neglected your duty of care, so you will be charged with criminal liability. The idea of the judge is the same, and the duty to avoid consequences comes first. Consider whether or not there is a consequence avoidance obligation, and how much it is, and make a conclusion first. In addition, in many cases, in legal theory, it is necessary to say that there was no predictability in order to acquit, so it is said that there was no predictability, and in order to convict, it is necessary to say that there was predictability, so it is said that there was predictability, and this is an afterthought.

Therefore, I would like you to learn the Road Traffic Law in which the duty to avoid consequences is written, and make a program that faithfully follows it. Specific examples of the Road Traffic Law that are likely to be problematic are the first sentence of Article 38, Paragraph 1, Article 36, Article 42, and Article 70. Many of the obligations stipulated in the Road Traffic Law are actually covered to some extent by the current automated driving, but it is these first sentence of Article 38, Paragraph 1, Article 36, Article 42, and Article 70 that are difficult to cover. By the way, apart from adults, what we really need to protect are elementary and junior high school students. You must not expect elementary and junior high school students to obey the road traffic law, and you must think they will not. Mothers often say that boys are like going to die, but I myself thought I was faster than a car when I was little, about 4 or 5 years old. So, there was a case that my mother intentionally let go of the hand she was holding and ran out in front of the car. I was almost run over, but the car stopped suddenly and my life was saved, and I thought the car was faster after all. Boys in elementary and junior high school do such unexpected things that much. So, if they make it on the premise of observing the road traffic law, there will be an accident. We have to make sure that the driver of the car and the person who makes the program of the automated driving follow the traffic law properly. The most accidents among elementary and junior high school students are at crosswalks without traffic lights and when turning right or left at intersections with traffic lights. The first sentence of Article 38 (1) covers these two points. And the other is that there are no traffic lights or pedestrian crossings in the first place (including T-junctions), and it is said that there are accidents at crossing points, which are Articles 36 and 42, but many accidents occur here. The last seventy is a bit foreign, so I'll talk about it again later. Many adults don't protect the crosswalk very well, but children do. I just defend, but I jump out. Therefore, an accident occurs. The regulation to avoid it is the first sentence of Article 38, Paragraph 1. We talked about this in detail at the fourth meeting of the sub-working group last time.

So what should we do to apply this to automated driving? I think that the way of writing the provisions such as "except when it is clear that there is no person" in the first sentence of Article 38, Paragraph 1 will be difficult when creating a program for automated driving. Therefore, I think that it is necessary to write down the specific cases that do not fall under the "case where it is clear that there is no person" in the specific government ordinances and rules. The blue part of the material is quoted from the Road Traffic Law Counselor introduced from the last time. When a pedestrian, etc. is standing at the entrance of a crosswalk, etc., but it is unclear whether it is waiting for a car to pass, waiting for a person, or trying to cross, when there are obstacles such as a telephone booth and a fence at the entrance of a crosswalk, etc., and it is unclear whether there is a person or not, when there is a street tree in the center of the road and the right side of the crosswalk, etc., is not clearly visible, when there is no street light on a rainy night, or when a crosswalk, etc. is installed at the end of a curve with poor visibility, etc., you must drive at a speed at which you can stop.

The pictures on the right and left of the document are in the same place. It is a crosswalk with no traffic lights, and you must proceed at a speed at which you can stop. There is a hedge on the right side, which is when it is not clear whether there are people or not. Look closely at the picture on the right. There is a white wall of concrete under the brown hedge on the right, but the front wheel of the stroller that I put on purpose is photographed a little behind the other side of the wall. The person who pulls the stroller is naturally behind, and the stroller is positioned in front, so only the front wheels of the stroller are slightly visible. In such a case, it is not obvious that there are no people. Then, step forward a little at a time until you reach just before the stop line, and you will finally see the whole picture of the stroller. You don't know until you come before the stop line. Another photo is of the road in front of the elementary school. There is a crosswalk at the end of the curve, but because it is curved, you can't see the left side at all unless you go near the stop line. If I were you, I would hold the government responsible for defects related to the installation of crosswalks in accordance with Article 2 of the State Redress Act. We shouldn't make a crosswalk in a place like this, but actually, there are many crosswalks like this. Even in these situations, you must drive at a speed that allows you to stop just before the stop line by natural braking without sudden braking. I think it is not enough to say whether such a thing will be programmed in automated driving now.

There are five levels of automated driving. At level 1, you can take your feet off the accelerator and brake pedals, at level 2, you can take your hands off the steering wheel, but you must pay close attention to the front, and at level 3, you don't have to look ahead. Cases where people are riding and crossing at level 2 are now operating at specific locations throughout Japan. At level 2, it is almost a automated driving because it is a foot-off and hand-off. If something happens on the way, a person will suddenly take over. I have conducted an undercover investigation on a course where both cars of this level 2 automated driving are running. Unfortunately, the automated driving vehicles did not slow down at all even when it came just before the. Then, the driver noticed it and switched to manual operation for the first time, but it was too late at that time and the car passed in front of me without slowing down. I did not mention the name of the company because it would be necessary to identify the company, but there is actually such a level 2 automated driving vehicles running. automated driving vehicles

Next is the problem and solution of Articles 36 and 42 of the Road Traffic Act. In fact, even among lawyers, there are some who view this article as a problem. To begin with, in the case of a normal crossroad without crosswalks or traffic signals, without exception, if the visibility is poor in any case, you must always slow down. Even if the lane in which you are driving is more than three or four times wider than the intersecting road, you must slow down unless it is a priority road. As a requirement, it says that at intersections with poor visibility on the left and right, without traffic control, and including not only crossroads but also T-junctions, you must slow down regardless of the width of the road and the intersecting road, except when the course in which your vehicle is traveling is a priority road. However, this is a little too strict, so as a practical treatment, in reality, when the road on which your vehicle is traveling is clearly wider than the intersecting road, even if an accident occurs without slowing down, prosecution will be suspended, or a fine will be imposed even if he or she is prosecuted. In reality, the driver on the intersecting road side is responsible, and it is as if he or she is not responsible. Moreover, the meaning of "clearly wider" is applicable if it is more than two times wider than the intersecting road, and it is not applicable if it is less than 1.5 times wider than the intersecting road, but the current situation is that the court cases are divided between 1.5 times and 2 times. Please take the next page.

I had a hard time taking these photos because of the heavy traffic in this photo location. It is about three times as high as Otsuma Street near Digital Agency. The pictures on the right and left are different locations. In the picture on the left, the ratio of the road width between the traveling lane and the left intersection road (T-junction) was 24 to 18, or 1.33 times, as I measured. So I can't say "obviously broad." Therefore, there is an obligation to slow down, but no one was slowing down. I don't think there will be any motivation to drive slowly. In the cross road (T-junction) in the picture on the right, it is 24 to 15, or 1.6 times, which is exactly where the court case is divided. It means that they don't know whether they are obliged to go slowly or not. When it comes to this, I think there is only one way to make automated driving into a implementation. In places where it is difficult to know whether there is a priority relationship or whether it is wide or narrow, one of the roads is uniformly made a priority road. There are two ways to display the priority road. It is a way to put up a triangular road sign and make it a priority road, or to write priority road on the road surface. However, there are few of these in Japan. If you look at the picture, the center line does not penetrate the intersection, and there is no center line in the first place. There is no traffic lane line either. This is exactly the indication that it is not a priority road. Priority roads are indicated when the center line runs through the intersection or when the first and second lanes of the traffic lane run through the intersection. By the way, even if it passes through the outer line, it is not a priority road. Then, I came to the conclusion that it would be impossible for automated driving to become a implementation no matter how much technology developed, unless the priority road relationship was always decided uniformly by drawing the center line in one direction at all intersections with poor visibility where traffic is not controlled.

Finally, I would like to talk about Article 70. This is also a provision that I think is impossible to clear. It says that the driver of a vehicle, etc. must operate the steering wheel, brake, and other devices of the vehicle, etc. without fail, and drive at a speed and in a manner that will not cause harm to others in accordance with the conditions of the road, traffic, and the vehicle, etc. It just states the obvious, but the reason why this is written is that it is only left as the last treasured sword of the family in the case that the individual result other than Article 70 cannot be covered by the duty of avoidance and duty of care. When this is applied, it is quite difficult when driving a car. Typically, if there are pedestrians, they must be at least 1 meter away, but if they are bicycles, there is no specific regulation on how many meters they must be away. It is often used in such cases. To be honest, I do not know how to solve this problem. I would like to limit myself to raising the issue that there is a need to further deepen the discussion at this point in the future on whether or not there is no other way but to revise the Road Traffic Act. That is all. automated driving

Mgr. Kozuka: Thank you very much, Takahashi. I think the discussion will advance very much as you raised issues based on specific situations. Now, I would like to move on to Agenda 2 - (3). The Secretariat will explain the main opinions of the 4th Sub-Working Group last time and the expected issues based on those opinions. Thank you very much, Counselor Suga.

Counselor Suga:
(Hereinafter, explanation will be made based on "Material 8: Fifth Secretariat Explanatory Material")
Document 8: Based on the explanatory material of the 5th Secretariat, we explained the main opinions at the 4th SWG, expected issues based on the opinions at the 4th SWG, the direction of the compilation, and the future schedule.

Mgr. Kozuka: Thank you very much,

Mr. Imai: Thank you very much, . I'm sorry I couldn't answer your question last time. As the chief examiner said, the case mentioned in (2) last time is a typical case that shows the direction to expand the scope of punishment for negligent offenders. As a premise for this, even in the case of Kaikei-byou, the basic stance of the court is to recognize the intention of the deceased even if he / she did not intend to kill him / her if he / she acted with the intention of killing someone and someone died. There was an incident in which the accused tried to kill a police officer and steal his pistol, but when he fired at the police officer, the bullet penetrated the officer's shoulder, refracted and flew, hitting two pedestrians more than 20 meters away and seriously injuring them. At that time, the Supreme Court admitted an attempt of robbery and murder against the two. As the defendant's lawyer, he argued that since he could not see a pedestrian who was more than 20 meters away in the dark, he was merely found to have been negligent and not intentional. However, since he fired a pistol in an attempt to kill a police officer, he had a very bad will, and the Criminal Code does not protect a specific person, but about everyone, so he showed his understanding that it is fine to recognize the defendant's intention even in relation to a person who the defendant did not recognize. The theory at that time supported this precedent, but the assertion that it was unreasonable was increasing since I was a student. At present, at the academic level, this precedent is wrong and is said to be a "Coelacanth precedent" to be broken. It is called the Coelacanth case because it represents a medieval understanding, an understanding since Roman law that since a bad will has begun an action, all consequences must be borne. Since negligence is a psychological state that is less than intentional, for example, if it is recognized that the person sitting next to the driver was careless, such as when he could have prevented the injury of the person but could not imagine doing so, I think it is the position of the judicial precedent to divert the negligence to the person who was sitting behind him who was not visible to him, and to recognize the negligence of the person who was actually injured. In this sense, the point made by the chief inspector is correct. When the case (ii) was issued, the discussion was quite lively, and the abstract legal code theory of willful misconduct itself was unreasonable, but it was criticized that whether or not negligence would be expanded according to that idea. However, recently, the number of young researchers who are aware of this point has been decreasing, but this abstract statutory code understanding seems to be contrary to the principle of responsibility. Based on these precedents, and considering what the Chief Investigator said, there is something in common with the opinions of Dr. Takahashi and Dr. Inadani. With regard to the question of the extent to which the possibility of predictability or the possibility of avoiding the consequences of an accident that was not expected at the time of creating the program can be questioned, in addition to my previous remarks, I believe that the possibility of predictability as negligence is denied in relation to a person who invested funds to prevent the occurrence of the consequences of infringement of legal benefits and was able to recognize that fact. This is the idea of measuring predictability in terms of the possibility of avoiding consequences. According to this understanding, if the design was made based on various materials at the time of the design, for example, knowing that the possibility of accidents occurring on this road is low and that the possibility of accidents occurring is this level, it is not possible to immediately confirm the predictability even if the possibility of exceeding it occurs several years later. However, for example, if there was a circumstance that the situation gradually changed while a car was driving on a road with ODD set, and traffic violations occurred frequently, as a manufacturer who knew it, I think that the duty to revise the program is based on the duty to act, and the predictability is confirmed when the new situation is reached, that is, there is a case where a negligent offense as an crime of omission is established. In the previous discussion, I failed to make a statement, so I added something. That's all.

Mgr. Kozuka: Thank you very much, . Regardless of what the theory thinks, in general, the current way of thinking adopted by the court is applied to cases of negligence, in particular, and furthermore, it is not a situation where you are actually driving a car and watching it in front of you, but is applied to negligence at the stage of programming at a previous point, so I think the court will probably operate it in a way that suits it. On the other hand, I think you pointed out that that and updating depending on the situation are different issues. Mr. Imai, thank you very much.

I would like to ask the other side to make a statement. In relation to the data part that was mentioned in the Secretariat's explanation earlier, I would like to ask Mr. Kuki of Honda Motor Co., Ltd. to speak as a guest speaker about what kind of data can be collected. Since it is an automobile manufacturer, could Mr. Hatano introduce it to us?

Member of the Hatano Committee: Thank you, Thank you very much. Regarding the relationship between the data, we are continuously exchanging opinions with the Subworking Secretariat. As part of this, it has been talked about what kind of practical examples there are in Japan, so we will ask Mr. Kuki to provide information from Honda Motor Co., Ltd., which is already on the market.

Mgr. Kozuka: Thank you very much, Thank you very much. Then, Mr. Kuki, nice to meet you.

Kuki clan:
(The following explanation is based on "Material 7: Summary of Honda Legend's Data Records")
Thank you very much. This time, I would like to explain about the data-recording installed in the Honda LEGEND Traffic Jam Pilot. My name is Hiroyuki Kuki, Electronic Control Unit development Section, Electronic Platform development Department, Honda Motor Co., Ltd.. Nice to meet you. Let me briefly introduce myself. I was in charge of the ADAS development of Honda LEGEND, which was launched in 2021. Please turn to the next page.

This is the outline of the automatic operation device. First, it is the external recognition around the vehicle. There are two cameras in the red circle in front, five radars in the blue circle in front, side, and rear, and five riders in the green circle in front, side, and rear. In addition, it is equipped with a high-precision map and a global positioning satellite system called GNSS for the recognition of the location of the vehicle. There is a driver monitoring camera for the detection of the driver's state, which detects whether the driver is facing forward or whether the driver is closing his eyes. Functional redundancy will be described later. As necessary responses and equipment for the automatic operation device, there are cybersecurity, software update, operation status recording device, which will be described later, and an outward indication sticker indicating that this car is equipped with an automatic operation device. Please go to the next page.

This page is the redundant design part of the system I mentioned earlier. The functions of the automated driving devices are roughly divided into five. First, at the top, as the lane recognition and object recognition devices, redundant design is performed by two sets of lidar and camera and radar and camera. In addition, as the control and instruction device, the main ECU is normally operated, and if an abnormality occurs in the main ECU, redundant design is performed so that the sub-ECU performs control. The speed reducer function is a redundant design of ESB and VSA. The steering function appears to be one of the EPS, which is a steering device, but the motor and electric-related parts inside are provided with redundant design. The power supply function is also provided with redundant design, including DC/DC converter and battery, and complete dual redundant design. Please go to the next page.

It conforms to the Safety Standards enforced in April 2020. On the left side of the document, it is the type approval necessary for the approval of general vehicles, but in this case, there are some additional items added to the automated driving vehicles. This time, I would like to explain the operating condition recording device. In addition, on the right side of the document, it is stated that the traveling environmental condition and the cybersecurity are required for vehicles with an automatic operation device. I will also explain the contents of the traveling environmental condition later. Please go to the next page.

The main driving conditions for the function called our Traffic Jam Pilot, which provides automated driving in the event of traffic congestion on highway main lines, are described on this page. As for road conditions, we provide it on expressways such as expressways and metropolitan expressways, or on dedicated roads for automobiles. However, we do not provide it in areas where the lane is not structurally separated from the oncoming lane by a median strip, etc., or in conditions such as sharp curves. In addition, as for environmental conditions, we do not provide it in situations where the lane is not visible due to heavy rain, snowfall, etc., or in situations where visibility is extremely poor, which is impossible for even ordinary drivers to see. In addition, as for traffic conditions, we provide it only in situations where there is traffic congestion or in situations close to traffic congestion. In addition, as for driving conditions, it starts to operate at a speed of less than 30 kilometers per hour and ends when the speed is 50 kilometers per hour or more. In addition, it is provided under driving environment conditions in which high-precision map information is available, the driver is wearing a seat belt, and the accelerator, brake, steering wheel, etc. are not operated.

The next page is an overview of the operation status recording device. History and time series data are stored in design units ① and ② in the middle. Toward that unit, information on recording timing is circulating from the SRS airbag on the right. Each unit receives vehicle information and sensor information from other control units and stores it. If one unit malfunctions or is broken due to a collision, the other unit stores it.

The following page contains the specific contents to be saved. I will not read out all of them, but as the status of the automatic operation device being operated, the status of the start of operation, the status of non-operation, the operation request to the driver, etc. are saved. In addition, the cause, time, time, location information, etc. when the above information was saved, as described in the upper right corner, are also saved. The software version is not saved, but it can be specified from the update history. As described below, as the information other than DSSAD, the behavior of the car such as acceleration, deceleration, steering angle, etc. are saved, and the video of the outside of the car when the automated driving is operating is also saved. However, since this is outside DSSAD, retention is not guaranteed. Please go to the next page.

This is a video as an example. As you can see in the video, regarding the information from the multi-view camera around the car, we store videos of the front, rear, and side. In our case, there have been no accidents, so the video is for normal driving conditions, but this is an example of the video. There is also a personal data, so the recording is at a resolution that does not allow you to see the license plates of the cars around you. Please go to the next page.

The information on this page is explained to the customer at the time of contract. It is used with the consent of the customer because it is equipped with a function to record data on the operation and control of the vehicle, the operation history installed in the vehicle, and so on, and as shown in the third line from the bottom, camera images around the vehicle while driving on the highway main line and position information of the vehicle are recorded. That's all from me. Thank you very much.

Mgr. Kozuka: Thank you very much,

Inadani Member: Thank you very much, Mr. . Now, using the materials I submitted in advance, I would like to comment on the materials compiled by the Secretariat. First of all, I would like to express my deep gratitude for the excellent compilation of the complicated and specialized discussions so far, and for the completion of the draft report, which shows steady progress in the discussions. In my materials, from the perspective of further complementing the discussions so far, which show such wonderful progress, I stated the necessity and ideal form of an effective corporate sanction system to prevent companies that manufacture automated driving vehicles, particularly overseas companies, from hiding organization's performance and data.

In the first slide, we briefly summarize the discussion on liability law so far. We understand that the essence of the discussion so far is to provide an incentive for automated driving vehicles manufacturers to manufacture automated driving vehicles with a certain probability of occurrence of accidents or accidents below a certain probability by understanding the reasons for liability occurrence such as "defect," "failure," and "fault" in connection with safety standards in type certification, etc., which are updated based on accidents or attempted accidents information.

The next slide shows the problem of the asymmetry between the burden of proof and the information, which is an inevitable problem when such a mechanism is adopted. In other words, there is a structural information gap between the side that is trying to prove the cause of the liability and the automated driving vehicles manufacturer. Therefore, if there is no way for the liability pursuer to access the information of the automated driving vehicles manufacturer, the liability mechanism may not function well.
As shown in the slide on page 4, this problem is known as the "adverse effect" problem of the Liability Act in the context of corporate scandals. From the viewpoint of a company that has information, if disadvantageous information is handed over to a party pursuing liability, there is no incentive to provide or disclose information because there is a risk of collateral damage such as damage to reputation or administrative disposition, as well as direct legal liability such as compensation for damage. Therefore, it has been pointed out that if legal liability or social criticism becomes heavy, companies will take action such as hiding information in order to evade responsibility, and as a result, there is a risk that the liability system will not function, which is an adverse effect. In fact, in the United States, despite the existence of grand juries with strong investigative powers and an extremely heavy corporate sanctions system, it is said that the corporate sanctions system did not function effectively until the introduction of the Deferred Prosecution Agreement system, which is said to have played a major role in resolving the "adverse effect" problem.

As shown in the slide on page 5, there are concerns about this "counterproductive" issue in the discussions of the Sub-Working Group. For example, last time, as one of the solutions to this problem, I understand that there was a reference to the establishment of a relationship between an insurance company and an automobile manufacturer in the scene of claiming compensation under the Automobile Liability Act. However, the incentives between the two are fundamentally different. If the cost of claiming for the collection of evidence is high, it may be reasonable for the insurance company not to claim for the damage. Therefore, it must be said that it is unclear whether the "adverse effect" problem can be solved by voluntary cooperation. Furthermore, we believe that such problems may become more serious, especially in the case of foreign tech companies, with which it is difficult to build cooperative relationships in the first place. On the other hand, since the investigative authorities that pursue criminal responsibility have the power of compulsory investigation, it is expected that this "adverse effect" problem will be solved to a certain extent. However, it can be said that it will be quite difficult to solve the problem if a huge company falsifies data or conceals or destroys evidence in a organization. I must say that this problem will become more serious if it is caused by a foreign company whose extraterritorial application is problematic. The direction of establishing the Independent Investigation Commission and strengthening its authority, which is stated in this report, is an extremely important step in resolving the issue of "adverse effects," and if it is realized, it will undoubtedly be a major step forward. However, bearing in mind the risk of the organization concealment activities of foreign companies that I have just mentioned, I believe that it will be possible to achieve greater effects by developing the legal system in a way that can deal effectively with the "counterproductive" problem. As you are aware, the Japan Fair Trade Commission and others are currently exploring ways to effectively apply the Japanese legal system to foreign tech companies that tend to run on a profit-first policy, such as the development of a Japanese version of DMA. Even in light of this situation, I believe that addressing the issue of "adverse effects" is a very important issue, at least in the medium to long term.

The slides from page 6 consist of three pages and summarize the basic ways to deal with the "counterproductive" problem. These are based on an analysis of the Deferred Prosecution Agreement System, which is said to have a track record of resolving the "counterproductive" problem, from legal and economic perspectives. The first slide summarizes the introduction of the surcharge system, which is gradually increased according to the viciousness of the company's behavior, and the whistleblower reward system, in order to ensure that it is more economically beneficial for the company to act honestly.

The slide on page 7 summarizes the expansion of damage due to "damage to reputation," which is a matter of concern for companies, the implementation of structural reform plans to eliminate questions and concerns of consumers and citizens, and the ideal form of "reorganization" of companies in which public institutions are involved, such as the appointment of independent supervisors to supervise them.

In the slide on page 8, it has been empirically observed that in order to effectively implement extraterritorial application, it is necessary to hold the "market" hostage, and that when a system that is persuasive from an international perspective spreads, international cooperation and market "unification" progress, and the effect of extraterritorial application increases. Therefore, these points are summarized.
The solutions presented so far were compiled based on the experience of the Deferred Prosecution Agreement organization, particularly from the perspective of effectively addressing regional illegal acts of globally active companies through the extraterritorial application of domestic law. We believe that the solutions will be very useful in complementing the discussions of the Sub-Working Group this time, particularly from the perspective of establishing a more effective accident investigation system, making the liability system function more effectively, and continuing efforts to steadily learn from failures and distribute safer automated driving vehicles.

The last slide is an attempt to discuss a legal system that embodies the above solution. I have said a lot, but I would be grateful if I could contribute a little to the further progress of this sub-working group. Thank you very much.

Mgr. Kozuka: Thank you very much, . Then, Mr. Takahashi will raise his hand from the venue.

Takahashi Member: I have been talking about criminal matters so far, but I would like to ask one or two questions about civil matters. Mr. Harada's slide 3 says that the current Road Traffic Law has "vehicles" as the subject in many of the traffic rules, and the obligations of "drivers" and "users" are set. Does it mean that it is rewritten as "users" in the case of automated driving?

Member: In the case of a complete automated driving, it is necessary to reorganize the obligations of the conventional driver into the obligations of someone. The addressee is likely to be the user or the manufacturer, and the intent is to allocate to one or the other.

Takahashi Member: user mean the holder or owner of the car?

Member: It will be like that.

Takahashi Member: In that case, I feel that automated driving does not have a sense of satisfaction because he is driving a car without permission. What do you think?

Member: You are right. I have shown that there are two theoretical possibilities.

Takahashi Member: I understand.
Next, I would like to ask about Dr. Inadani's presentation. In the part where the problem is located, it is stated that if the victims, insurance companies, or prosecutors cannot prove the existence of "defects," "disabilities," or "faults," the manufacturers of automated driving vehicles will not have an incentive to produce safe automated driving vehicles. I think you are right from the viewpoint of incentives. In fact, under the current law, the prosecutor has the burden of proof in any case, so I don't think there is a problem with this point. The problem is the victim or the insurance company. In the sense that the insurance company makes a claim, the victim makes a claim to the insurance company, the insurance company pays the claim, and the insurance company makes the claim, but under the current law, I understand that the insurance company has the burden of proof when the insurance company makes a claim. I think the problem is that if the victim cannot prove the claim, the victim will not receive any money and there will be no relief. I think it would be better for the victim to change the burden of proof and for the insurance company and the prosecutors to bear the burden of proof as before. What do you think?

Inadani Member: Thank you very much, Mr. . In the first place, the purpose of my report is that insurance companies and prosecutors may also be in an extremely difficult situation. In particular, I have just mentioned that it will be considerably difficult if organization concealment acts are carried out. This will lead to the problem of how to solve the problem that even if the burden of proof is shifted in the case of the victim, the company will disprove the fact that it has been shifted, and the materials that appear in the disproof are only those that are convenient. Therefore, shifting the burden of proof is a measure that can be extremely important, but in addition to that, it has been empirically observed that foreign tech companies, in particular, will fight against the hiding of information and evidence. So, I am saying how to think about this point.

Mgr. Kozuka: Thank you very much, My understanding of the material is that what is written as "if we cannot prove it" is not an incentive for manufacturers "if we cannot prove it." It is actually the same as what Dr. Takahashi said, and Dr. Takahashi said that the burden of proof should be shifted as a way to eliminate it, but I think Dr. Inadani's intention today is that we must consider various things including that.

Takahashi Member: , I would like to introduce my example. I filed a medical care error lawsuit about 20 years ago, but I did not submit my medical records at that time. As a result of various measures such as establishing penalties for that, there is no medical institutions that does not submit medical records now. In addition, electronic medical records cannot be falsified. If we create such a system in automated driving, we will have to submit information, and I think it will be difficult to falsify.

Mgr. Kozuka: Thank you very much, . I would like the Secretariat to consider where to take this, and is there anyone who has any opinions on any other points in the presentation? Then, I would like to ask Mr. Hatano and Mr. Ochiai in that order.

Member of the Hatano Committee: Thank you, . I would like to confirm a little about the burden of proof that was just discussed. Regarding automated driving, as information has already been provided by Honda Motor Co., Ltd. today, due to the revision of the Road Traffic Act in 2020 or the revision of the Road Transport Vehicle Act, the record of the operating state of the automated driving device, which was not required for conventional control devices, has been added as a requirement. Considering this, it can be understood from the industry that the prerequisite for proof has already advanced compared to the past by requiring such data by investigative authorities. Considering this, I think it can be understood that the essential issue of the burden of proof that is being discussed today is not a debate targeting automated driving, but a debate on the burden of proof for defects in automated driving, or in automobiles, as they are called. Considering this, it is possible that such a debate should be held not only for automated driving, and it is a little strange that only automated driving, which is moving forward with respect to conventional products, will be the subject of debate. Therefore, I would rather like you to try to hold a debate from a comprehensive perspective.

Second, with regard to the direction in which the flow was added, which was newly presented by the Secretariat, I would like to express my gratitude because I could sense the overall flow of the discussion so far and the flow was very organized, and I could feel that you made efforts. In addition, I would like to ask you one thing. In the column "Occurrence of Accidents and Incidents" in the flow chart, individual micro-accident investigations and the collection of near-misses, etc., are described in the context of the utilization of data. On the next page, (2) is organized in the form of a combination of cause analysis and data utilization. If you think about this before and after the flow chart, it would be easier to understand if you could organize ⑤ and ⑥ into a group centered on accident investigations, and ⑦, ⑧ and ⑨ into a group centered on data utilization activities. I would like you to consider this. In addition, regarding ⑨, while I understand that the usefulness of infrastructure is an important task, I think that it is still difficult to step into utilization just by confirming its effectiveness. Therefore, if its effectiveness is confirmed, I would like to say that it will lead to the further use of ITS if you consider the development of systems and systems related to infrastructure certification, management, and operation, if possible. That is all from me.

Mgr. Kozuka: Thank you very much, . All of these points are important, but regarding how to read the materials, is it correct to understand that (5) and (6) are about individual accident investigations, and (7) to (9) are about joint use of accident and near-miss information?

Secretariat: Since the squares are divided, I have summarized them in (2), but the upper two and the lower three are completely different in nature. The upper one is originally called micro, and the lower one is called macro, so I would like to devise it by adding a middle heading or dividing the boxes. In addition, I received a talk about infrastructure. I would like to consider what I can do with the ministries and agencies in charge, perhaps through demonstration.

Mgr. Kozuka: Thank you very much, . Mr. Ochiai, Mr. Imai is raising his hand, so I would like to hear your opinion. Then, Mr. Ochiai, thank you very much.

Ochiai Member: Thank you very much. I would like to make a few points.

My first question is about the matter that Dr. Inadani and Dr. Takahashi discussed earlier, which is to ensure that information is disclosed. In terms of example measures, I believe that the basic response is to include the obligation to cooperate in the investigation of (5). In medical institutions, I think there are parts that are similar to what was done by administrative regulation for the disclosure of medical records, so I think there are parts that are being asked to take measures here. I think it is very important to have information provided in order to solve the case properly, so I think it is important to maintain it properly. In addition, I believe that the sanction systems that Mr. Inadani mentioned should be discussed head-on here, but in relation to ride-sharing, there is also talk about the regulation of white taxis. For foreign corporations, for example, the fact that they have not been registered as a corporation in Japan or that they have not appointed a representative has been a problem in the Travel Agency Law and the Telecommunications Business Law. If anything, I think there is a possibility that overseas operators will directly come in and provide it, so it is necessary to consider these points more. In addition, in terms of the principle of conduct at Big Tech companies, I believe that there are cases in which it is acceptable to ignore sanctions if they are weak. In relation to the Telecommunications Business Act, for example, I am aware that there are carriers who have overstepped the non-criminal fine for failure to register, so for example, if it is several hundred thousand or several million, I think it will be ineffective. For current Japanese manufacturers, I feel that it is sufficient if ⑤ is maintained, but based on other matters, I think that Dr. Inadani's discussion will come out.

The second point is that since I participated in the meeting in the middle of the day, I saw the materials. On page 4 of Mr. Harada's materials, you pointed out that it was Issue in automated driving, and I believe that you pointed out the update and immediate application of security standards and the administrative responsibility for updating security standards. Regarding this update part, for example, in terms of examples of measures, I think there is also a part that has been organized in consideration of a mechanism to require compliance with the Safety Standards (10). On the other hand, what I think is important is that the certification standards or guidelines are appropriately updated when the social environment or technology changes, which I think is especially behind the part that says administrative responsibility in this comment. The mechanism for this is basically an accident investigation system for accident liability, a macro information collection system, and an arrangement of safety standards, so I feel that the necessary functions are secured to some extent as a component. However, regardless of whether or not it is written in the figure, from the perspective of agile governance discussed by the Ministry of Economy, Trade and Industry, for example, once information is collected, I think it would be good if it could be changed in a big loop, such as updating safety standards and guidelines. In the end, I think that it will go to the court and there will be a dispute over whether the Safety Standards and Guidelines were appropriate standards. However, I think that the fact that we have a system to collect information in a timely manner and review it when necessary will not only ensure the appropriateness of the content but also help maintain the final legal stability. I would like you to consider this.

The third point is what Mr. Hatano said earlier, that there may be a part that is related to the whole. Certainly, there may be a part that is related to the whole. On the other hand, as a slightly different part, I think there is a part about how to evaluate overseas corporations and the part that puts more emphasis on the software part. On the other hand, as it is often discussed at the Review Meeting on Digital Related System Reform, which is being discussed in Digital Agency, there is a discussion that equal footing is important when there are actually only things and when digital is combined. As I said earlier, there may be different elements, but I think it is a reasonable opinion to reflect the same elements in the existing ones as much as possible.

Fourth, you pointed out the utility of infrastructure. I believe that how to maintain it and how to determine the roles of each party that will be the premise of the responsible relationship will be extremely important external restrictions when actually making a development. I would appreciate if you could summarize this point in (ix), including where to entrust the infrastructure. That's all from me.

Mgr. Kozuka: Thank you very much, . There were various points of discussion, but I would like to ask Mr. Imai to make a statement. Thank you very much.

Mr. Imai: Thank you very much, , I believe that today's report by Dr. Harada and Dr. Takahashi is of great significance for the future compilation. I had some questions about it, so I would like to ask you two some simple questions.
First of all, in Mr. Harada's report, the word "driver" is used, and you say that there will be no drivers when the automated driving becomes more advanced, but how do you understand the concept of the driver? Probably, from the viewpoint of administrative law, the entity responsible under administrative law is different from the entity responsible under criminal law, so the concept of the driver may be different depending on it. As a premise, there is no definition in the Road Traffic Law, so I would like you to tell me how you thought and wrote it like this.

In relation to Mr. Takahashi's explanation, I basically think in the same way, but I thought it would be necessary to comment a little more on the point that it is the Road Traffic Law that specifically states the obligation to avoid consequences. The Road Traffic Law is a regulation to protect road traffic safety in general and in the abstract, and as the Ai No Kai said today, your main concern is what to do in the event of a violation of life or human physiology, and it may be wrong to leave this to the Road Traffic Law. Therefore, based on the points I reported last time on the crime of negligence in the pursuit of social activities causing death or injury, negligence or the obligation to base negligence should be considered, and in that case, the violation of the Road Traffic Law will be an important fact, but it will be only indirect evidence. With that idea, I would like to confirm whether there is any inconsistency with Mr. Takahashi's opinion.

Mgr. Kozuka: Thank you very much, , I would like to ask you to answer each of your specific points. First of all, Mr. Harada, please tell us about the concept of the driver.

Member: Thank you for your question. As you said, I think the concept of a driver under the Administrative Law and the concept of a driver under the Criminal Law are different. Here, it is assumed to be a concept under the Administrative Law. In other words, in relation to the fact that the Road Traffic Law sets various action obligations, the term driver is used here to mean a person who is engaged in activities that enable him / her to fulfill his / her action obligations, that is, a person who actually operates the steering wheel. That is all.

Mgr. Kozuka: Thank you very much, . Also, I would like to ask Mr. Takahashi about the relationship between criminal negligence and traffic law.

Takahashi Member: : To simplify the discussion, I just said road traffic law, and I think it is more accurate to say that the original way of writing is road traffic law and other related law. For example, I think there is a T-junction where you turn left, which I mentioned earlier, and the road traffic law says that you must slow down there, but recently the Public Prosecutor's Office has begun to say that in some cases, there is a temporary stop duty not only for slowing down, but for example, in the case of large trucks such as tank trucks. This is not stated in the road traffic law. This is said in the form of a duty of reason, but if you want to write in detail about automated driving, I think you must write about such things and related law. That is all.

Mr. Imai: Thank you very much, . I understand Mr. Takahashi's purpose very well. That is still an important point, but no matter how quickly the Road Traffic Act is revised, there will be some parts that cannot be written down. It is similar to what Mr. Ochiai said earlier, but I came here because I thought that by making thorough preparations when setting ODD, at least in relation to automated driving, such a problem could be avoided. In addition, on the other hand, if other people misunderstand Mr. Takahashi's opinion, it may be that there is no fault if you follow law. As there have been recently, there have been completely misguided claims that automated driving is not at fault if you follow ISO, so I would like to say that we will not be able to send a message here. Thank you very much.

Mgr. Kozuka: Thank you very much, . Mr. Inadani is also raising his hand, so Mr. Inadani, please.

Inadani Member: Thank you very much, Mr. . I would like to express my thoughts on the questions Mr. Hatano asked. Mr. Ochiai has already pointed out several points, but I think there are three main points about automated driving that need to be considered.

First of all, I think it can be pointed out that cheat acts such as data fraud are extremely sophisticated and difficult to detect when software is used. In the case of Volkswagen, the method was quite sophisticated, and it is not difficult to imagine that such a thing would be a serious problem in the context of using AI, which is more advanced software.

In addition, as Mr. Ochiai said, I think we must consider the fact that the behavioral principle of tech companies is quite aggressive compared to conventional companies. The second point is that it cannot be denied that there will be parts that must be considered based on a considerable theory of evil.
Third, there is a possibility that automated driving vehicles will eventually become a digital platform. If so, a network effect will occur, and it will have a significant impact on a considerably wide range. I think it can be pointed out that this is a point that must be feared compared to a standalone automobile that does not use software including AI.
In response to the above points, from the perspective of protecting Japanese companies that are serious about protecting the law, as well as from the perspective of protecting the trust and safety of consumers and citizens, I think that if we do not think about what we can handle properly, it will be very bad. Therefore, I have made various comments today. Thank you very much.

Mgr. Kozuka: Thank you very much, . Is there anyone else who would like to speak? Mr. Takahashi is raising his hand from the venue. Then, Mr. Takahashi, please.

Takahashi Member: Secretariat, it is very difficult to say, but shouldn't Article 2 of the State Redress Act be specified in the civil liability? If the update of the security standards is delayed, the responsibility for the national compensation will come out after all. Dr. Harada talked about the national compensation earlier, and there was a similar problem in the Chikuho Pneumoconiosis Lawsuit. In the steel field, the standard was set so that damage would not occur, such as draining water, but only coal was left. Because it had been left for several years, the national compensation was approved. In the update of the security standards, accidents occur every day in automated driving, so I think there will be situations where the responsibility for Article 2 of the State Redress Act will be asked in a short span of time. I would like to include room for that.

Mgr. Kozuka: Thank you very much, . The point you just pointed out is not to consider infrastructure and the National Compensation Law again, but to update the security standards and guidelines in (10). If this is delayed, the national government may be responsible. Since it may be agile governance as Mr. Ochiai said, I think it is better to write with a creative way of thinking rather than increasing the number of action items. Since you are writing a written report, I would like you to write it for sure in it. Is there anyone else who would like to speak? Then, Mr. Takahashi, thank you.

Takahashi Member: I heard that it will be documented, but when will it be?

Mgr. Kozuka: Thank you very much, Is it correct to understand that we will adopt the report at the next Sub-Working Group?

Secretariat: Within a month, we will be incorporating our writing work, consultations with each ministry, and consultations with teachers, so I am very sorry for the short notice, but I would like to make efforts so that it will not be seen for the first time at the next meeting.

Mgr. Kozuka: Thank you very much, I believe that we will receive your opinions at the next meeting.
Then, as I have time, before I make a summary statement as the Chief Examiner, I would like to step down from the position of the Chief Examiner and say a few words about the fact that I am the only civil law professor who is absent today. We just talked about the burden of proof. When we consider the burden of proof in civil cases, the claim from the accident victim is a tort. Of course, there are special laws such as the Self-Compensation Act and the Product Liability Act, but these are also illegal acts. Even if the insurance company makes a claim, it will be subrogation in the end. In that case, when an accident occurs and damage is caused, I don't think it will work as a court system to ask the manufacturer and the people concerned to prove who was at fault. The burden of proof in the legal sense probably remains on the plaintiff, and I think it is necessary to develop a solid system so that it will not be difficult to pursue responsibility in the problems surrounding the burden of proof, such as the formation of points of contention and the submission of evidence in the process. I understand that what has been discussed since a short while ago is the responsibility for verification activities in a broad sense. If we extend it, there is a discussion about how to fundamentally consider the operator's liability under the current Automobile Liability Act itself, but it is necessary to consider it again in the medium to long term among the action items this time, so I think the understanding is that we will proceed with the current system for the time being, based on a report compiled by the Ministry of Land, Infrastructure, Transport and Tourism in 2018, and that will be sufficient for the time being.

From here on, I would like to wrap up today's meeting from the perspective of the Chief Inspector. First of all, thank you very much for providing various suggestions today. I received many points from Mr. Harada, but first of all, he pointed out that there are parts of regulation related to current road traffic that depend on criminal liability, and I think this meeting has clarified the background why criminal liability has been a major issue until now. Second, regarding the issue of safety standards, he pointed out the importance of updating and immediate application in particular. This is because the Building Standards Act, etc., makes existing ineligibility a principle, but considering automobile accidents, it is probably difficult to make it a principle, and I understand that it will be sorted out in the future examination of specific safety standards.

I think it is very important that Mr. Ai no Kai also made a valuable point and said that he is hopeful that the progress of automated driving will fundamentally lead to traffic safety and, in turn, to the prevention of unfortunate and sad people in society as much as possible. In the Secretariat's report, the magnitude of automobile liability for the accidents, or conversely, expectations for automated driving, are also stated. I think this point is extremely important, and I am not sure if I can say it from my position, but I would like to see manufacturers make development that meets such expectations.

In addition, Mr. Takahashi talked about the necessary system, and in addition to that, Mr. Takahashi talked about the incorporation of rules on road traffic. I think this has become very clear in the last few discussions, so I think it is important how to implement this in the future.

I would also like to address the new points you pointed out today. First of all, there were discussions on the application of the Road Traffic Act to overseas business operators on several occasions today. I believe this is a major issue when considering the current market situation, so I would like you to take this into account as much as possible in your report. Today, there was also the issue of enforcement, but I also wonder if we can respond to the issue of incorporating the Act into the programming of the Act. In other words, we can have close dialogue between Japanese manufacturers and the Government of Japan on how to incorporate the Road Traffic Act into the program, so I would like to ask you to do so. However, if we do not do the same thing to foreign manufacturers, I believe it is necessary to consider the issue of equal footing in industry and the safety of the Japanese people so that foreign manufacturers do not produce products that do not incorporate the Road Traffic Act at the same cost as Japanese manufacturers and claim that they meet the safety standards themselves in that state. In relation to this, Mr. Ochiai and Mr. Inadani pointed out the ideal form of sanctions. There are major issues, so I would like you to take this into account as well, although it may be a medium - to long-term issue for Issue.

Another point is that you pointed out that the frequency of updates is a criminal responsibility or an administrative responsibility, and it is ultimately related to state compensation. I would like you to take this point seriously in your report. That is what I have considered from today's discussion.
In addition, regarding the sharing of data, Mr. Kuki of Honda Motor Co., Ltd. told us about it in detail. Based on this, I would like to ask you to summarize this subworking at the end. Mr. Hasui, Vice-Minister for International Affairs, thank you very much.

Director Hasui: I am Hasui from Digital Agency, . Once again, I would like to thank you all for taking the time to gather today. In addition, I would like to apologize for the malfunction of the digital environment. In the five discussions so far, including today, I would like to thank you very much for teaching me the significance of the points of contention and for the very specific points such as Mr. Takahashi's talk on the road traffic law. I feel that the resolution is increasing steadily. I have written a flow chart as a low-resolution version. I have received various opinions on it today, so I would like to consult with you in advance about how much I can summarize it, including those opinions, by switching on and off during the Golden Week holidays, as I was told by the Secretariat earlier. After that, I would like to make arrangements with related ministries and agencies. I look forward to your continued support. I have already received a summary from you, so I would like to once again express my gratitude and request for your kind attention. Thank you very much.

Mgr. Kozuka: Thank you very much, . I would also like to thank you for your valuable opinions today. If you have any additional opinions, please send them to the Secretariat by next Friday. In particular, I have heard that some members will be absent because they are overseas today, so please give us your opinions. As usual, we will publish the materials for today's meeting on the Digital Agency website at a later date. The secretariat has told us that it may take a little more time than usual. In addition, as usual, we will publish the minutes on the Digital Agency website after confirming them with the members.

The next Sub-Working Group is scheduled to be held in late May. The agenda is currently under consideration, but at least the draft report that was mentioned earlier will be released. I will contact you later, so thank you very much.

With that said, I would like to close today's 5th Meeting of the Sub-Working Group on the Study of Social Rules for automated driving vehicles in age of AI. Thank you all for taking the time to attend. Have a good weekend.