Skip to main content

This page has been translated using TexTra by NICT. Please note that the translation may not be completely accurate.
If you find any mistranslations, we appreciate your feedback on the "Request form for improving the automatic translation ".

The 6th Meeting of the Sub-Working Group on Social Rules for automated driving vehicles in age of AI

Overview

  • Date and time: Thursday, May 23, 2024 from 1:30 p.m. to 3:30 p.m.
  • Location: Online
  • Agenda:
    1. Opening
    2. Proceedings
      1. Explanation by the Secretariat (Main opinions of the 5th age of AI Sub-Working Group on the Review of Social Rules for automated driving vehicles, main contents of the report (draft), and future schedule)
    3. Exchange of opinions
    4. Adjournment

Materials

Minutes

Councilor Kodama: Thank you very much, Mr. First of all, I would like to make an administrative contact. Today's meeting will be held completely online. Members are requested to turn on the camera during the meeting and to unmute the microphone when they speak. If anyone else is speaking, please mute them. In addition, spectators are requested to turn off the camera and microphone.

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 secretariat explanatory materials, the report (draft), 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 email. 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 have heard that member Goto will participate in the middle of the meeting, and member Ochiai, member Sakamaki, member Suda, member Nishinari, and member Yoshikai will leave the meeting in the middle of the meeting.

This time, as in the previous meeting, Mr. Ohno and Mr. Suganuma from DADC are also participating as guests. Please note 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.

Mr. Kozuka: Thank you very much, Mr. . We will proceed according to the agenda. First, I would like to explain about Agenda 2 - (1) Secretariat. Mr. Suga, Counselor, nice to meet you.

Counselor Suga:
(Hereinafter, explanation will be made based on "Material 3: Explanatory Material for the 6th Meeting of the Secretariat")
Material 3: Based on the materials of the 6th Secretariat, we explained the main contents of the report (draft) and the future schedule.

Mr. Kozuka: Thank you very much, Mr. . We would like to hear the members' opinions on this, but prior to that, we have the materials that Mr. Hatano has submitted. At the second meeting, you said that it is necessary to take time to consider the cooperation of data, and we have held repeated discussions with the secretariat, so we would like to report on the results this time. Thank you, Mr. Hatano.

Member of Hatano: Thank you for choosing Thank you for the opportunity to introduce us. As the Chief Inspector introduced earlier, since the second meeting, the JAMA and the Sub-Working Group Secretariat have been carefully discussing data utilization in parallel. Since this is the sixth and final meeting, we would like to share the results of the discussion with you in the form of a discussion and use it as a reference for future discussions.

This material is a technical consideration on the collection of near-miss data from the viewpoint of preventing accidents and improving the driving environment. As you know, regarding accidents that could not be predicted in advance, including those outside automated driving, you said that it is desirable to collect some other data and then analyze it objectively. However, if data is collected at random, it is difficult to analyze and analyze it. In short, there is a concern that a large amount of unnecessary data will remain, and important data will be unexpectedly missed. Therefore, it is necessary to consider the need to collect data that is firmly useful as evidence. This time, I think that there are mainly two areas in the recording of operating data. The first is the data necessary to analyze and analyze the cause of an accident. As introduced by Honda Motor Co., Ltd. last time, there are two aspects. One is that accident analysis can be performed in accordance with the law by collecting the minimum data with legal binding force. On the other hand, there are two aspects. If near-miss data, which does not directly lead to an accident, is captured to some extent, it may be useful for accident prevention and environmental improvement. This time, I especially focused on the second near-miss data.

First, we tentatively determined that a near miss refers to a scene in which a condition that does not result in an accident but may deviate from a predetermined safety margin is observed. If you try to acquire this kind of information, there are several possible Issue. It is not limited to near misses, but since it is a motion recording, it is necessary to determine in advance the minimum data set required to correctly analyze and analyze the target scene, which is a combination of data including the type, number, and recording time of the data, after determining the trigger condition to start recording the target scene in the first place, in other words, the trigger to start recording. It is also possible that the trigger condition does not hold. In other words, we are faced with a scene that cannot be recorded in advance, but unfortunately, it is difficult to record such a scene because the trigger does not occur. With this understanding, it is also important to collect data after confirming what it will be useful for. In addition, the data recorded may be specific to a combination of system and environment in which you are driving. In other words, since it is possible to measure only the car and the scene, it is necessary to carefully handle the expectation that the same thing will be measured in other systems. When obtaining near-miss data, we believe it is necessary to carefully organize it from the viewpoint of the automated driving system. Please take the next page.

Based on this premise, I will briefly consider how to actually collect data with illustrations. Let's take a case study where a automated driving vehicles is following a car in front of it. At first, I keep the following distance, but at ②, the car in front of me starts braking suddenly for some reason, and the following distance is blocked, but automated driving vehicles has not responded yet. At ③, the distance between the cars is too close at last, and the brake is generated. This becomes the trigger condition, and if both parties continue to decelerate, the car will eventually stop. The main point is whether or not they finally collide, but I think that the necessary data will be recorded from such work as a series of flow. In this case, if we do not define in advance what the difference is between a normal scene and a near miss, we may record only the data of a normal scene. If we do not set up the trigger to be recorded before and after the trigger, we may not know what happened even if we record only the moment of the trigger. If the trigger occurs when the vehicle in front is absent, it may be a different event, or as I mentioned earlier, it may occur only in the vehicle in question, so it should be noted that it may not be the same for other vehicles. If we try to extract several specific near-miss scenes while taking these points into consideration, there are cases where there is a sudden change in acceleration / deceleration, cases where there is an unexpected approach of neighboring vehicles, cases where the vehicle deviates from the scheduled traveling road, and cases where there are several physical quantities that can be triggers, scenes to be assumed, and data to be recorded at the same time. It is important to pay attention to and make design in advance to record these things in connection with each scene.

The next page is a summary of what I said in the form of illustrations, so I will omit the details, but some patterns can be thought of in this way. If we statistically collect this kind of data, for example, as shown in the example on the lower left, in places where a large number of scenes were measured in which pedestrians entered crosswalks and automated driving vehicles suddenly braked, I think it will lead to the acquisition of evidence for improving driving environments, such as helping to make decisions on installing traffic lights for crossing pedestrians. By collecting this kind of data, we believe that it can be used to expand overall safety and social acceptability.
Today is just a reference case, so I think you will discuss more details in the efforts shown by the sub-working group this time in the future, but I would like you to use it as a reference. That's all.

Mr. Kozuka: Thank you very much, Mr. . Including this, I would like to ask one question. Do you mean that the trigger is the data about the car in question, and therefore it will be decided later whether it is a scene to be collected or not?

Member of Hatano: Thank you for choosing Both can be considered, but I think it would be desirable to collect them after narrowing down what kind of scene this trigger is supposed to be in advance if possible. If you simply try to analyze after the fact, there are cases where the data required for the trigger to be recorded at the same time is not prepared, and it is possible that the trigger will not be in a state that contributes to the analysis, so I think it is important to assume to some extent in advance.

Mr. Kozuka: Thank you very much, Mr. . I just wanted to confirm what you said. I think you said that the scene is also fixed, so it can be automatically extracted, and the necessary information collected can be extracted. Thank you very much.

Then, I would like to use the remaining time from here to ask for your opinions on the method of compilation, the content of the report (draft), the data collection you explained, and so on. Please press the raise hand button on Teams or show it in a chat. Some teachers have limited time, so please let me know when the time is approaching. Mr. Ochiai has already raised his hand, but Mr. Yoshikai has limited time until half past two in the afternoon, so could you change the order and start with Mr. Yoshikai?

Yoshikai Member: Thank you very much. I would like to give you two opinions about the report (draft) just in case.

I would like to point out one point. On page 17 of the draft report, I believe that the system of deferred prosecution agreements in the United States is introduced. As far as I know, the practice of prosecution in the United States is significantly different from that in Japan, and even in cases where there is a low likelihood of conviction, there are cases in which prosecution is carried out and settled by bringing the case to plea bargaining. In addition, even if a deferred prosecution agreement is reached, companies continue to commit violations of the law even after a compliance program has been formulated. There is an evaluation that deferred prosecution acts do not help deter corporate crime. In discussing the system of deferred prosecution agreements, I believe that it is necessary to carefully consider the differences in practical practices between Japan and the United States and the negative aspects of the system of deferred prosecution agreements. I would like to say this here, even if it is not worth writing in the report.

Another point is, as I believe it will be around pages 23 and 24 of the report (draft), as arranged in the four quadrants, even if an incident occurs due to automated driving vehicles, I believe that whether or not there is criminal liability is determined based on specific predictability and whether or not there is a violation of the duty to avoid consequences, as has been arranged so far. However, in actual accidents, I believe that whether or not to file a criminal prosecution is determined based on more specific facts while based on general and abstract standards. In the future, when specifying security standards and guidelines, for example, as I believe Mr. Takahashi pointed out today, in order to clarify standards such as how many km/h a straight car should travel when automated driving vehicles turns right, I believe that it is necessary to analyze the responses to individual and specific accidents for which criminal liability has been imposed so far, mainly based on published court cases and the like. If we attempt to quantify security standards or guidelines in the future, I believe that it is difficult to quantify them without referring to court cases and the like. By doing so, I believe that it will be possible to verbalize specific standards and make the Road Traffic Act machine-readable. That is all.

Mr. Kozuka: Thank you very much, Mr. . I think these are all important points. I would like to reflect them in the report.

Then, could you speak to Mr. Ochiai?

Ochiai Member: . I would like to make a statement on the materials you are projecting now. I created the materials I submitted this time because I thought it would be better to clarify the relationship between data connections, the investigation of the incident, and the update of the rules. In the discussions in the Sub-Working Group so far, in the part shown in the figure, I touched on the "Update of Security Standards and Guidelines" at the lower left. I believe that there will be discussions in the future on how the design of the investigation will be conducted. However, I believe that it is important to reflect the results of the collection and analysis of information on accidents and near-misses at the macro level in the security standards and, in some cases, in the traffic rules themselves. At this time, it may be necessary to update not only the safety standards and guidelines but also the traffic rules one by one. In doing so, it is necessary to take into account the examples of judicial decisions and administrative dispositions. In addition, I believe that appropriate rules will be formulated by thoroughly collecting not only accident and near-miss information but also technical information described in the macro section and updating them in terms of respective safety standards and rules. On the other hand, regarding the collection of accident information, there may be manufacturers and operators. However, since it is too detailed, the operators are not shown in the figure, but I think that how to collect information from infrastructure providers is also important. It is written in green as Issue 10. In addition, I think it is necessary to have a process in which this macro information collection is fed back to the manufacturers. It is described as Issue 9. I believe such a process is also necessary, and through such a mechanism, we should advance cooperation and collection of information in development, as well as the rules. Of course, the micro of individual matters will also be organized, but as stated in the report (draft), it is still necessary to organize the relationship with the investigative authorities in the future, and there is still debate over how to disclose and use the results of the investigation. Therefore, I think that the arrow may originally come out from the left side, but at present, there are some parts that are under consideration, so they are not described. In fact, in the "data connections" on the right side of the figure, I think there is an issue of what kind of information linkages will be made between the infrastructure provider and the manufacturer, and how to consider the demarcation of responsibility there. In addition, I think the issue of what to do with the information linkages in the relationship between the manufacturer and the insurance company, which is described with * in Issue 12, is also an issue. If funds, etc. are included in Issue 12, there is a possibility that we will consider that, but if we go that far, we will make assumptions one after another, so I have not specified it. I believe that it is important for appropriate general system development to make it possible to see the relationship between each issue of information collection, update, and design sophistication so that cooperation can be made. Therefore, I have prepared and presented these drawings. That's all.

Mr. Kozuka: Thank you very much, Mr. . This is also a very important point. Looking at the report (draft), I feel that it is a little weak in terms of how to use data collection, etc., so I would like to reflect this.

Next, Mr. Imai, nice to meet you.

Member Thank you very much. First of all, I would like to ask a simple question about Mr. Hatano's report today. Can I assume that the trigger conditions will be updated in the course of the investigation? Probably, the term near-miss is used when there are many unexpected accidents, so if the trigger conditions are fixed, I feel that there will be many oversights from an amateur's point of view, so I would like to ask a simple question.

Mr. Kozuka: Thank you very much, Mr. , what do you think?

Member of Hatano: Thank you for choosing

Member . Now, I would like to give you my opinion briefly. You have done a great job in compiling the report (draft). I am impressed that it has been compiled so well in such a short period of time. However, I would like to mention only the points because I think there will be future Issue.

First of all, as a point of concern while reading, I think that the description of automated driving vehicles is written on the premise that there is no driver, but there was a way of writing automated driving vehicles in which there is an equivalent to a driver. As far as I understand, the description is academically correct, but since it may cause a misunderstanding of the reader, I think that it is better to unify it, or if it is shared here that there is actually a case where there is a driver even in automated driving vehicles, it is better to supplement it later.

Second, regarding the behavior of the automated driving vehicles, several requests are written, and it is stated that the first is to observe the Road Traffic Law. Then, ③ and ④ will come up as cases that cannot be dealt with. I did not have much time to discuss ③ and ④, but I thought that it would be better to write that the principle of observing the Road Traffic Law is a principle, and that there has been a discussion about in what cases exceptions should be allowed, because the danger cannot be avoided only by observing the Road Traffic Law.

Third, with regard to negligence, there have been discussions on the comparison of civil and criminal negligence, and I believe that Member Yoshikai made a related remark earlier. I think that I think in the same way as Member Yoshikai, and the court shows judgment on a case-by-case basis, so I do not think that the precedents, for example, focus on the obligation to avoid consequences. As I said in my report, I believe that the court or the prosecutor who will prosecute is probably considering the presence or absence of negligence by looking at the predictability and the possibility of avoiding consequences in an integrated manner, so I think that it is necessary to summarize the results after giving due consideration to such points.

Finally, I believe that Ochiai members provided a very easy-to-understand explanation on the ideal form of the investigation agency. As I believe you mentioned in your explanation, the debate on what the investigation agency can say to the subject of the investigation, the issue of the duty of the subject of the investigation, is quite sensitive and has not yet been fully discussed. And, for example, I think that the issue of the duty cannot be solved unless we weigh the utility that can be obtained if the duty is imposed on the other party against the possibility that information will be obtained if the subject is made to speak at will. I believe that it will probably be Issue in the near future, so if there are many opinions like this today, I think that it should be reflected in the report. I believe that the duty has not been agreed here. I hope that you will write about it carefully. That is all.

Mr. Kozuka: Thank you very much, Mr. . We received many important opinions and suggestions. Next, Mr. Takahashi, I would like to ask for your cooperation.

Member Takahashi: It is at the end of page 27 of the Please project my resume. Up to now, I have mentioned three problematic points in the creation of safety standards. They are Article 38, Paragraph 1, Article 36, Article 42, and Article 70. These are as I have explained.
Another point is the problem of Article 34. This was pointed out by a member after the end of the previous sub-working group, and I thought that Article 34 would also be a serious problem. In this Article 34, the specific method of turning right or left is written. On page 3, it is written that when turning left, you should move to the left side of the road and slow down, and when turning right, you should move to the center of the road and turn right inward. In fact, accidents do not occur very often regardless of whether you follow it or not. The problem is that it is written in red, "What is actually problematic in criminal liability is not the violation of the method of turning right or left itself, but a right-hand accident with an oncoming straight vehicle, and to what extent should a right-hand turning vehicle consider the excessive speed of a straight vehicle." This is a point that is quite controversial.
As you can see on the next page, the court precedents are all over the place. Of course, it is necessary to assume that a car going straight will exceed the legal maximum speed or speed limit by 10-20km/h, but the problem is that the court precedents are divided when the speed is 30-40km/h. It is necessary to assume that the speed exceeds 30-40km/h, and it is not necessary to assume that the speed exceeds 30-40km/h. The difference is, for example, in the Tokyo High Court in C, it is said that "it is not necessary to assume that the speed exceeds 20km/h," but I feel that it is much lower than the court precedent that 30 to 40km/h should be assumed. However, looking at this specific case, it is only said that it is not necessary to assume that the speed exceeds 20km/h because it is impossible to assume that the motorcycle is running on the side strip, which is a sidewalk. As you can see, it is all over the place depending on the specific road environments. The problem is how much should be assumed when converting it into a automated driving. Please show me the next page.

Court cases vary from case to case depending on traffic environments. However, I think that the fact that it should be assumed that the speed exceeds 30km/h at least is generally the majority of many court cases. However, in order to make automated driving possible, the Road Traffic Act will be revised, and unless it is uniformly written that it is assumed to be 40km/h just in case, I think that it cannot be read as automated driving vehicles when programming. What we must be particularly careful about is that in a right-handed accident, a car going straight increases its speed in a hurry because it becomes a yellow signal, or increases its speed because it is approaching the intersection just before it becomes a red signal. Therefore, it is necessary to program uniformly as much as possible while considering how many km/h is assumed before how many meters from the intersection. This is the first point.

One more thing, I would like to talk about civil matters. I would like you to take a look at page 25 of the report (draft). In the third paragraph, it describes the Self-Compensation Act. The burden of proof is naturally transferred to the Self-Compensation Act, and the perpetrator is responsible. Please take the next page. In the third paragraph of 1.3.3, with regard to civil liability, "In this case, a person who is a victim who files a claim for damages (excluding the case of pursuing the operator's liability under the Self-Compensation Act) shall be punished by the court if he / she falls under any of the following items: In the claim based on Article 709 of the Civil Law, it is necessary for the insurance company that makes the claim to assert and prove the causal relation between the fact that the automated driving vehicles did not conform to the security standards, etc. and the damage (that the incident could have been prevented if the airport had conformed to the security standards, etc.). " In the actual trial, there is almost no problem if the operator and the offending driver are the same owner car, but for example, if an employee is driving a company's car, the employee is the offending driver and the company is the operator. The company has the burden of proof transferred under the Liability Act as an operator and is liable in almost any case. On the other hand, since the offending driver is subject to Article 709 of the Civil Code, it is argued that the offending driver does not have to be liable for damages unless the victim proves it, but this is not the case. In actual practice, the complaint and the judgment are written in parallel, "The victim seeks compensation for damage of several yen from the offending driver based on Article 709 of the Civil Code, and compensation for damage of several yen from the operator based on Article 3 of the Compensation Act." The court's decision is to largely ignore the burden of proof under Section 709 and focus solely on the liability law. Since there is almost nothing that can be disproved in Article 3 of the Compensation Act, we will assume liability. At that time, the victimized driver who is sued based on Article 709 of the Civil Code is actually considered to be responsible in the same way as the Self-Compensation Act without considering the burden of proof. This is the practical operation. In the report (draft), Article 709 is written with great emphasis, and when it comes to automated driving, Article 709 seems to be quite visible, but I think there is a problem here. In fact, there are no cases where the burden of proof is on the victim side under Article 709, and all cases are filed under the Self-Compensation Act. Both the wrongdoer and the operator are in fact doing it under the Automobile Liability Insurance Law. As for why such a thing is allowed, I think this is a little different from a murder case, for example. If it's a murder case, you may be a victim once in a while, but you can hardly assume that you will be a perpetrator. However, the world of cars is different, and everyone becomes a victim or perpetrator. All 120 million people would be victims, and if half of them had a driver's license, any of the 60 million people could be perpetrators. So there is such compatibility. In such a case, if an accident occurs, we should compensate at least with money, and let's solve it with money for the time being without thinking about the difficult matter of the burden of proof. Therefore, there is a part like the spirit of social solidarity and mutual assistance in which everyone pays insurance money and pools together to help each other. This is the basis of today's automobile society. Because of this system, everyone can drive with peace of mind and walk outside with anxiety. When this becomes automated driving, if Article 709 of the Civil Law comes out and it becomes apparent that the victims have the burden of proof, they will probably be too scared to walk outside and will not be able to own a car. If such a thing is common in the world, social trust in automated driving should also be lost. So, I have been talking mainly about criminal liability, but I think there are still various problems in civil cases. Here, I would like to emphasize that we would like you to maintain the idea of Article 3 of the Self-Compensation Act. What I am a little worried about on page 25 of the report (draft) is that it says, "We will also consider the concept of operator responsibility for Level 5 owner cars, taking into account the progress of technology and the status of new business models." I can understand this examination if it is an examination of who will be the operator, but I am opposed to it if it includes the conversion of the burden of proof, that is, returning the burden of proof to the victim side. I would like that part to be more clearly stated in the report. That's all.

Mr. Kozuka: Thank you very much, Mr.

Next, Mr. Inadani, please.

Inadani Member: . First of all, I would like to once again thank the Secretariat for carefully summarizing the complicated discussions. For myself, this sub-working group was really a lot of learning, so I would also like to thank the participating teachers for their lively discussions.

In this talk, I understand that the aim is to update safety standards and related laws and regulations, including road traffic laws, based on data, to update social systems in the sense of changing the road environment itself, as pointed out by Mr. Hatano, and to pass liability legislation in connection with these. In addition, by operating liability legislation using survey results based on scientific knowledge at that time, the legal system and social system themselves will be updated according to the development and spread of science and technology, and the legal system will be built to appropriately control risks. This is a very important step toward the realization of a new governance mechanism called agile governance, which has been noticed and approved by the WEF, the OECD, the G7, and others. I believe that we have made tremendous progress in realizing responsible innovation without excessive atrophy. On the other hand, as various professors have said, I think there are still some points to be considered. If I take up some of them, for example, there are path-dependent problems, but I think there are some difficulties related to the way of operation and interpretation of the Liability Act. This may be related to what Dr. Takahashi mentioned earlier, but from my understanding, I think that until now, we have been discussing how to grasp the causal relationship in a deterministic manner by targeting the occurrence of individual and specific results, and how to handle the negligence related to that. The issue at issue this time is rather a type of discussion on how to suppress potential risks throughout society, and I think that the bad alignment here will inevitably remain. This will lead to a review of the legal use of the burden of proof and the subject of proof, as well as consideration of the ideal form of liability legislation itself. I believe that we, as an academic society, need to continue to discuss this issue, and I would like to contribute to this effort.

I believe that automation of road infrastructure will be discussed in the future. Or, I do not think that the problem of automated driving vehicles, which is a complex system called a system of systems, which is what to do when multiple systems cooperating with risk management are involved, was directly dealt with this time. However, Mr. Ochiai pointed out earlier in this meeting that there will be a problem of responsibility sharing when infrastructure is automated. If such a problem occurs, the management method this time is a kind of centralized method, and the government is responsible for creating safety standards. If this method is used, as Mr. Harada pointed out, the problem of sharing the responsibility of the provider of services and products and the responsibility of the government of regulation, which controls with responsibility, may be the way of the multi-stakeholder process in the update of regulation, but I think that such a matter also needs to be further organized.
In addition, if the system becomes extremely complex, I think there is a possibility that we will have to aim for more autonomous distribution in the future. In that case, with regard to the review of the Liability Act, a transition from liability without fault to full strict liability may come into view. In considering a mechanism to more strictly ensure the authenticity of data and information used to maintain security, that is, a mechanism that can autonomously improve security, I believe that it will be increasingly necessary to build a sanctions system that uses economic incentives for companies and a whistleblower protection and reward system to ensure the authenticity of data. In this regard, Mr. Yoshikai pointed out earlier that the agreement to postpone the prosecution may be problematic. I personally believe that it is necessary to discuss the issue based on data and evidence at this meeting. At least in terms of data, the number of exposed cases differs greatly between cases without such a system and cases with such a system. In addition, if there is a violation of the agreed provision after the agreement has been reached, sanctions will be imposed, and by receiving even stricter responses, illegal activities will be suppressed. Therefore, I believe it is necessary to make the system more effective by at least comparing it with the case where there is no such system, or by discussing the problems of the existing system based on data and evidence. As for the difference from business practices, I may be saying it a little too much, but in the field of economic crimes related to technology, there are various cases such as the Winnie Incident, the Coinhive Incident, the Pressance Incident, and the Okawara Kakoki Incident. Regarding whether or not the business practices of this part should be given, if the relationship between them is an issue, I think it can be said that it is okay to consider it from a wide range of perspectives.
Finally, I have said many things, but it is difficult to create a complete form that cannot be complained about in a short period of six months, and it is not desirable to be overly hasty about such an important issue. Therefore, I think it is important to practice agile governance, which is to continuously and rapidly review the ideal form of the legal system from a wide perspective based on data and evidence in response to the development and spread of science and technology in the remaining Issue or the Issue that is being discussed. I think that the discussions of the subworking this time focused on aiming for a better legal system based on data and evidence, and the report (draft) of the results also states such things, so I think that the subworking this time was a big step forward in the sense that such points were confirmed. That is all from me.

Mr. Kozuka: Thank you very much, Mr. . You have pointed out some important points.

Next, Mr. Suda, please.

SUDA Member: . Regarding the investigation of the accident that I was advocating, I think it was good that the conclusion was in a very good direction. However, there is a point that I was concerned about in the part of how to write it. The purpose of the investigation of the accident is basically to determine the cause and prevent recurrence, and the pursuit of responsibility is the work of the police, so I think it would be better to go a little further on that point. That is my first point.

In addition, when the Safety Standards and the Road Traffic Law are revised, it is important to update the software and hardware in response to them, but in reality, I think that updating the software and hardware requires technical consideration. It is not as easy to update as a computer or a smartphone, and it is necessary to make a validation to determine whether it is truly functioning correctly, and in some cases, trial operation is also required, so I think that in-depth discussion is necessary. I thought that there might be a discussion about who will update, for example, if a repair shop performs the update, the repair shop may be responsible, so I thought that consideration on this point will be important in the future.

Finally, in the middle of page 20 of the main text, at the end of Collection, Analysis, and Use of Accident and Incident Information (Macro), there is a description that "* The results of' careful discussion' will be reflected as necessary." I do not understand the meaning of this, so please tell me what it means. That is all.

Secretariat: Regarding the last point, "careful discussion" has become a proper noun in the Secretariat, so I am very sorry. It will be like an internal memo, but we call the story explained by Committee Member Hatano at the beginning "careful discussion", and we would like to reflect the essence of it, so I wrote it.

Mr. Kozuka: Thank you very much, Mr. Do you mean that what Mr. Hatano explained today will be included here?

Secretariat: Yes, I would like to add it.

SUDA Member: Okay.

Mr. Kozuka: Thank you very much, Mr. . Next, Mr. Imai and Mr. Ochiai would like to make a second speech. Due to time constraints, Mr. Sakamaki and Mr. Nishinari will be attending by 3 pm. May I have a speech from them? First, Mr. Sakamaki, please.

Member Sakamaki: Thank you very much, . I have read the entire report (draft), and there are no particular objections. In particular, in relation to my specialty, I think it was very good because I can predict that specific discussions will advance through short-term discussions in Issue regarding the investigation of the accident and the coordination of the investigation and the investigation. In addition, regarding the relationship between the investigation and the investigation, several people mentioned that they would require or obligate the parties concerned to report, but even in the existing accident investigation and future prevention committees such as the Japan Transport Safety Board, the provision of information and reporting from the parties concerned exists as a classic problem, and I recognize that it is an important Issue that should be continued in the future, and at the same time, that specific attempts should be made for coordination.
In relation to crime, criminal investigation, and criminal liability, which is my or Mr. Imai's specialty, investigation begins when the investigative authorities think that a crime has occurred in a certain fact. When considering whether or not there is a crime, in particular, where the fault is located because automated driving is almost culpable, the main premise common to investigation and investigation is to determine what the objective facts were and what the mechanism leading to the accident was. In the case of crime, it is about how to evaluate criminal liability, so regarding the common premise, first of all, it is necessary to have a system that can collect as accurate information as possible. From such a perspective, I think it is important to proceed with consideration of the relationship between investigation and criminal investigation, and to make it concrete and discuss it. That is all.

Mr. Kozuka: Thank you very much, Mr. . Then, I would like to ask for your cooperation.

Nishinari Member: Thank you very much, . I also have no particular objection, but I would like to make two brief comments.
One is that the system is currently complex, but there are many attacks around the world that try to break it with malicious intent. In that case, even if it is based on the guidelines in ODD, there are cases where accidents occur without knowing well. If a professional with such malicious intent comes to break the system, I think there will be a very difficult problem in the future. I think it is necessary to recognize the need for countermeasures once you experience it, and to update it with predictability. Therefore, I think we should consider countermeasures against malicious intent.
The other thing I think is important is location information by GPS of mobile phones. For example, even at an intersection where you can't see who is over there, there is a possibility that it is possible to detect the approach of a car and a person by using the location information. I thought that it would be good to discuss such safety services in the future. That's all.

Mr. Kozuka: Thank you very much, Mr. . You pointed out a very important point of view. Then, Mr. Imai, please come in.

Member .
The first point is a supplement to the point already made by Mr. Suda and Mr. Sakamaki. As Mr. Suda said, the purpose of the investigation is to determine the cause of the accident, and the investigative authorities are to hold accountable, but it is as Mr. Sakamaki said that the decision will be based on overlapping facts. This is a classic problem, and I have been considering various matters at the Antimonopoly Law Basic Problem Roundtable about 20 years ago, taking the opportunity of the Kawasaki-Daisya Incident. Even so, it is not easy to reach a conclusion, and we have been considering solutions that are good in terms of hasty action each time. As Mr. Sakamaki said, I think it is good to put in solutions according to needs through continuous discussion. In doing so, instead of requiring them to come out, we should always consider whether there is a system to give incentives to people to submit information, and I do not think that if we require them to come out, useful information will come out. I think Mr. Inadani said this point, and I think so too.

I have one more question. I would like to ask Mr. Takahashi. Thank you for your very meaningful report today. In relation to Article 34 of the Road Traffic Act, there was a suggestion about a right-handed accident. When there is a right-turning vehicle, how much speed should be considered in creating a program in relation to the opposing straight-going vehicle. I think Mr. Takahashi's suggestion was that a program should not be created based on the current situation, although it is probably a violation of the Road Traffic Act because straight-going vehicles often accelerate when they see yellow. That is a realistic solution, but if that is the case, it will be inconsistent with the main principle of automated driving vehicles observing the Road Traffic Act, such as Asimov's first principle, and I feel that automated driving vehicles will also be required to accelerate and turn right when entering an intersection. I think whether or not it is good should be determined by whether or not the number of accidents decreases statistically. Therefore, I think there are many points to agree with Mr. Takahashi's proposal, but we must consider something like the Asimov principle. So, I would like to ask your questions and opinions. If you have any comments, please let me know. That is all.

Mr. Kozuka: Thank you very much, Mr. , please.

Member Takahashi: It is at the end of page 27 of the I noticed it when you pointed out. If all cars are 100% automated driving, there is no way that oncoming cars will exceed the speed limit. So, I noticed that what I am talking about here is at the stage where 100% automated driving is not being used, and if about 30% is automated driving vehicles and 70% is not automated driving yet, it will be a problem. Thank you very much.

Member . It will be fine if everyone changes their behavior when they see automated driving vehicles's behavior, and it will also be a conversation about whether we have the tolerance to wait that long. I think we need to look at the situation in area, where ODD is set. However, at that time, I commented on the report (draft), but I feel that the part about how to evaluate de facto practices when determining negligence is not the same in civil and criminal cases. In criminal cases, everyone is crossing a red light, so even if I say that I also crossed the red light, I will be recognized as being at fault, so I would like to end my comment. Thank you very much.

Mr. Kozuka: Thank you very much, Mr.

Ochiai Member: Thank you very much. I have one big point and about three others.
First of all, I think the main point is that it is important to think about what the goal of the Study Group itself was and to continue to communicate it in the future. Overall, I think it is an initiative to make automated driving vehicles a social implementation by combining software well in order to reduce the number of traffic accident victims. I think it would be good if this point could be clearly conveyed as a message when it is announced at various places in the future.

As for the other three points, first of all, as Mr. Imai, Mr. Sakamaki, Mr. Inadani and others discussed earlier, the mandatory reporting on the investigation of the accident is important. I think that it is common that this point is subject to incentive design, and there are various ways of saying it, such as Mr. Inadani talking about the postponement of prosecution agreement system and the current system, but in any case, it does not necessarily work just to obligate, so I think it is important to devise how to create an effective form. Today, the method has not been finalized, so I think it would be good if you could think about what kind of innovation can be made at a relatively early timing. In that regard, I don't think it will be soon, but in the medium to long term, it may be a problem whether it will be an incentive design that can really be cooperated when overseas software manufacturers, and in some cases, Big Tech, are involved in the future, so I think it would be good if you could take note of such points.

The second point is about the machine-readability of the Road Traffic Act. In the draft report, there has already been a discussion that it is necessary to consider rules in some cases so that machines can be used, including the use of some physical means. I believe this point is extremely important. When discussing machine-readability, even if we are discussing it at the regulatory reform Conference or the Digital Dialogue, we tend to say that we should translate it as it is. However, if the rule description method is translated into a form suitable for digital processing, I think that it may be necessary to write it in a form slightly different from the original rule. I believe this point is important, and I believe that the same protection and legal benefits should be maintained, but I believe that it is not always a good way to write it even if it is rewritten as it is. I would appreciate it if you could keep this point in mind.

Finally, regarding the accident investigation agency and the update that Dr. Suda mentioned, I think it is very important. In particular, in relation to the accident investigation agency, in particular, regarding the micro point, there are newspaper reports and the like, and it is not necessarily done only for the investigation. I think it would be better to make it clear that it is separate from that, so that misunderstandings do not occur. In this summary, although the details should be discussed in the future, there are discussions that may cause misunderstanding, so I think it would be good if you could include an expression that would check that point. In fact, I think that setting that part firmly is the key to submitting materials and cooperating in the accident investigation, so I would like you to describe it so that unnecessary concerns do not arise. In addition, the update point that Dr. Suda mentioned is also a very important point, and I think that it will be a necessary point of contention when programs, programs, medical care equipment, and other AI and software are incorporated. In practice, I think it may be the most important as a safety measure, so in that sense, Dr. Suda's point is important. That is all.

Mr. Kozuka: Thank you very much, Mr. . Next, Mr. Sato, please proceed.

Member : Thank you very much. Dear members of the Secretariat, I am very grateful that you have compiled this long and wide-ranging draft report in a short period of time. I myself have made various comments, and I understand that most of them have been reflected, so I would like to make some comments, although I will not be discussing them specifically.

First of all, in relation to the government, in order to quickly and appropriately update safety standards and guidelines, I think it is important to collect as much useful information as possible. In relation to this, in the discussion of updating safety standards earlier, I made a few comments in footnote 57. As a premise for updating automobiles, for example, if there is an urgent need, it is necessary to forcibly update to some extent. On the contrary, it is dangerous to operate an automobile that has not been updated. Therefore, I think it will be important to consider how to organize this in the future.

In addition, in relation to civil affairs, there was originally a report on self-compensation in 2018, but I think it is wonderful that it is stated in detail that we will review it again based on various discussions this time, and that it will be compiled during the twenty twenty-five. I myself was involved in the compilation of the report as an official at the time, but as Mr. Takahashi said, I believe that we must not retreat from the prompt relief of the victims. In addition, this report also describes the full relief of the victims, so I believe that we must never give up.
Apart from self-compensation, it is indicated that discussions will also be advanced on matters related to the PL Law. I understand that in the EU and other countries, there is discussion about including some software using AI in products, but it is very important to have discussions in Japan as well. I think there are difficulties from the beginning, such as how much products will be included in the scope, so in conclusion, it may be possible to conclude that the current court case is sufficient, but I think it is very important to have discussions.

In terms of criminal matters, even if automated driving is introduced, liability for negligence itself will be maintained and judgment will be made depending on the case, and it was difficult for users and development operators to move. On the other hand, in this summary, it is clearly indicated that if there are security standards and guidelines that are appropriate and reasonable, they will be appropriately considered when deciding on disciplinary action. I think that is quite a step forward. In relation to this point, there is a description about the remote supervisor in footnote 53 of the report (draft), and in the course of the comments in the report (draft) itself, there was originally a comment to the effect that the carelessness of the remote supervisor itself is unlikely to be a direct cause of the accident. However, when I look at the current outcome, it has been deleted. I think that cases (I) and (ii) as assumed here are premised on cases in which the remote supervisor is at fault in the first place, so I think it is better to re-include the original main premise that the carelessness of the remote supervisor itself is unlikely to be a direct cause of the accident. As I mentioned earlier in relation to criminal matters, it will be clearer from the perspective of the parties concerned from which range the negligence is made, so I think it is better to say so.

I understand that we have made considerable progress in the investigation agency field this time. When I worked as an official in the past, I thought it would be difficult to discuss the matter of the Japan Transport Safety Board conducting an investigation into the automated driving vehicles Incident. However, I think that the fact that the Japan Transport Safety Board is exemplified in this report (draft) is itself a very significant progress. I would like to see the reports and other materials that are the results of the investigation utilized for updating security standards and guidelines, and I hope that they can be utilized well from the perspective of ensuring the effectiveness of civil claims. It's a little long, but that's all. Thank you very much.

Mr. Kozuka: Thank you very much, Mr. . Mr. Imai and Mr. Takahashi have requests to speak, but in relation to the overall process, I would like to ask you to say a few words to each of the teachers who have not yet spoken. In particular, civil affairs was mentioned in Mr. Sato's speech, so I would like to ask Mr. Goto and then Mr. Nakahara for their opinions.

Member: Thank you very much. I came here late today, so I can't follow the discussions in the first half, but I think the report (draft) really had to cover a wide range of areas in a short period of time. Thank you very much for summarizing it. I have nothing to ask about the report (draft) itself.
I believe that rather than drawing any conclusions this time, we have created a roadmap. Perhaps the most important thing is that civil liability was discussed, but rather than discussing civil and criminal liability separately, I think that the discussion was held with considerable awareness of what kind of image should be drawn as a whole, including administrative regulation. From now on, roles will be assigned to individual ministries and agencies, but I believe that there is a possibility that we will return to individual discussions. Therefore, I expect that follow-up will be conducted by taking the initiative in Digital Agency and other places, and that discussions will continue to be held in consideration of how to design the system as a whole so that it will not be stacked somewhere. Thank you very much.

Mr. Kozuka: Thank you very much, Mr. . Mr. Nakahara, please.

Mr. Nakahara: Thank you very much for your comments. I do not mean to ask you to reflect the perspective of civil liability in the report, but I would like to make three comments on matters related to Issue in the future. Thank you very much for your comments. Thank you very much for your comments. Thank you very much for your comments. Thank you very much for your comments.

First, assuming that appropriate and detailed safety standards or guidelines will be established in the future, I believe that the significance of this report (draft) is that it sets forth a basic viewpoint that appropriate additional civil liability will be realized by appropriately considering them when the court decides on the disposition. Needless to say, the success or failure of the concept depends on how much concrete and quantitative standards are presented, so I would like to expect progress in technical discussions on this.

Second, in relation to the point I just mentioned, as reserved in the draft report, I believe that the judgment on the imposition of civil liability does not take into consideration only compliance with the Safety Standards, etc. In addition, I believe that how to reflect that in the framework of the discipline of operator liability and product liability is also an unsolved problem. This is a rather legalistic problem, and I expect progress to be made, and I myself must work on it.

Third point, which is also derived from the point I just mentioned, is the request for appropriate liability imposition, which is a request on the part of the responsible party, but on the other hand, the main purpose of the whole system is to appropriately compensate for damage or provide relief to victims, and to build a socially desirable way of sharing damage. If the civil liability system does not function properly, I think a drastic change will be necessary. In that sense, I recognize that the option of damage compensation by the fund was proposed in the previous presentation by Mr. Harada, who took on the challenge of including the burden on the administration, but I feel that cross-sectional consideration should be made in the legal field, including the system design, such as whether to position it as a main method or as a supplementary method when the liability system has limits. In any case, I would like to express my deep gratitude to all of the members, including the secretariat and the chief inspector, who made a valuable compilation on the difficult Issue. That is all.

Mr. Kozuka: Thank you very much, Mr. . Mr. Harada, please go ahead.

Member: Thank you very much. Regarding the proposal you mentioned earlier, I proposed it while saying that it would probably be impossible, but I believe that it will not lead to a fundamental solution unless we consider such matters. I believe that the fact that this report (draft) has been compiled with such a wide range of contents even though it was a very short period of time is extremely significant, and I have no particular comment on the report (draft) itself. With regard to the Administrative Law, I recognize that there are many issues that have yet to be actualized, and I believe that the relationship between the responsibility system and behavioral control must be discussed by the entire academic community, and then discussions toward institutionalization must be matured. Thank you very much for this opportunity. That's all.

Mr. Kozuka: Thank you very much, Mr. :

Member: Thank you very much for . Thank you for allowing me to participate in the Sub-Working Group, which will be an important position for automated driving's social implementation. In addition, I would like to thank all the members of the Secretariat for compiling the report (draft) in such a short period of time. I think it is very desirable that the number of accidents will decrease significantly as social implementation progresses in the future, but on the other hand, I don't think it is possible to completely eliminate unexpected accidents. Therefore, the non-life insurance industry will continue to make efforts to ensure current or better relief for victims by functioning as an infrastructure in such cases.

Thank you for incorporating the materials we submitted into the report (draft). I would like to refer to Part ⑫ of the future priority measures because I believe that the insurance industry is mainly involved here. As Dr. Sato mentioned earlier, in the 2018 Ministry of Land, Infrastructure, Transport and Tourism report, I believe that we will consider self-compensation on the basis of maintaining the conventional operator responsibility. Once again, we would like to cooperate with the consideration so that a complete compensation system and prompt victim relief can be secured.

As mentioned in the materials, I also touched on some property damage. As for not only accidents resulting in injury or death but also accidents resulting in property damage, although the accidents may be large or small, I would like to ask once again that all accidents not be out of scope. As stated in the place marked with * below, I think that a certain amount of data related to automated driving will be required, so for compulsory automobile liability insurance, which is currently Level 3, we are considering building a mechanism in which JAMA and automobile manufacturers provide data to insurance companies in order to ensure the effectiveness of accident situation investigation or investigation of the cause of the accident based on it. We would like to continue to discuss this so that a similar cooperation system can be established even if it becomes Level 4 or higher. I would like to ask the Government of Japan to continue to consider the development of a system for overall accident situation investigation and investigation of the cause of the accident.

Mr. Kozuka: Thank you very much, Mr. . Mr. Fujita is absent today, so I think I have received your opinions. Mr. Takahashi, I have kept you waiting, thank you very much.

Member Takahashi: It is at the end of page 27 of the Secretariat, the white frame says criminal liability according to the case as before, and the blue frame says criminal liability according to the case. I think this means that even if the security standards are not met, the criminal liability according to the case may not be assumed, and even if the security standards are met, the criminal liability according to the case may be assumed. In that case, I think it would be better to state it a little more clearly in the report (draft). It is written in the top paragraph of page 22 of the report (draft) that if the performance conforms to the security standards, even if an accident occurs, it is assumed that it can be evaluated in principle that the Road Traffic Act has been observed and that the accident could not have been foreseen or avoided. If a lawyer reads it, he or she can understand that there are exceptions in principle, but I think it would be better to add words according to the flow chart, such as that criminal liability may be charged depending on the case. What do you think?
Secretariat: Thank you very much. I have no other intentions, so I would like to do so.

Mr. Kozuka: Thank you very much, Mr. :

Member Takahashi: It is at the end of page 27 of the If it is written, it is fine. In addition, regarding the Accident Investigation Commission, it is stated in the second paragraph of the micro part on page 19 of the report (draft) that cooperation with the accident investigation is mandatory in the case of certain accidents such as serious accidents. Does it have the nuance that it is mandatory under law?

Secretariat: vehicles, we had an image of conditioning at the approval stage, such as asking for cooperation in an investigation if a serious accident occurred.

Secretariat: is a person who has obtained the approval of the automatic operation device, so it is a system to request the cooperation of manufacturers, etc.

Member Takahashi: It is at the end of page 27 of the Member: I think this is what the manufacturer is most concerned about, but in fact, it is also what the victims are most concerned about. From the manufacturer's point of view, it is internal information of the company, so they don't want other companies to know about it, and from the victims' point of view, they want all the information to be disclosed so that they don't have the same victims. As I think I said in the previous installment, I think it would be fine to set up a certain degree of compulsory investigation authority in the Investigation Committee. In this case, the members of the Investigation Committee would be like special national government employees, but I think it is okay to put such things in perspective, but if we do it too strongly, the manufacturer's willingness to development will be lost, so I am wondering what to do about this point.

Mr. Kozuka: Thank you very much, Mr. 's point has a difficult problem. I will speak as a member, not as a chief investigator, but I myself am a member of the Consumer Safety Investigation Commission, an accident investigation agency. The Consumer Safety Investigation Commission has compulsory investigation authority based on the Consumer Safety Act, but in the 10 years since its establishment, the compulsory investigation authority has never been exercised, and it has been operating by collecting information on a voluntary basis. In this way, I think there is a way to solve the problem by maintaining a delicate balance in operation. However, broadly speaking, I think it is necessary to consider how to best proceed based on the idea of the incentive structure that Mr. Inadani always says when design the system. Is that okay?

Member Takahashi: It is at the end of page 27 of the Until now, what was questioned was external responsibility such as civil liability and criminal liability, but there is another responsibility called disciplinary action. It is an internal disposition of organization. Certainly, if you are subject to criminal disposition externally, you will also be subject to disciplinary action internally. I had the impression that there was one way to separate that and give incentives.

Mr. Kozuka: Thank you very much, Mr. . It is traditionally explained that disciplinary action based on the Mariners Act is a major incentive for marine accidents, so I think there is certainly such a way of thinking. Thank you for pointing it out. Then, Mr. Imai, please.

Member . I wanted to hear what Mr. Takahashi just said again, but due to time constraints, I would like to leave it to Mr. Inadani. I am also a member of the Investigation Committee on Commercial Motor Vehicle Accidents, but I have only voluntary investigative authority, so I am struggling during the investigation, and I am doing what I can to suit the feelings of the victims. From that experience, I can understand a considerable amount of the facts even without compulsory authority, and in terms of preventing accidents in the future, I can say that it is quite effective. Therefore, as the chairman said earlier, I think it may be possible to decide what authority to have and how to place undrawn swords. I will leave this to Mr. Inadani.

The other is footnote 53, which Mr. Sato mentioned earlier. I forgot to say earlier, but I think the description of this point will require considerable consideration in the future. In the case of (1), where the remote monitor did not take measures to terminate it, it is said to be Level 4, but in reality, it is a case where it has not reached Level 4. That means, in (1), the driver may actually be the remote monitor, and (2) is the real automated driving. Most people do not have such a recognition, so this is fine as it is, but probably, unless we combine the definition of the automated driving, the definition of the driver, and what can be done even if they are not the driver from a remote location, etc., we cannot reach a conclusion. In addition, as described here, the recognition of the causal relationship is different between civil and criminal cases because of the complex factors. In the case of civil cases, the causal relationship becomes a problem on the premise of joint tort. It is the same in criminal cases, but the point of the causal relationship is not clear. In fact, it is a very important part, so it will be necessary to consider it in the future. If we can reach an agreement here, I would like you to add that the conclusion has not been reached. I thought again after hearing Mr. Sato's statement, so I made a statement. That is all.

Mr. Kozuka: Thank you very much, Mr. . I heard that there will be a supplement from the office.

Secretariat: I am sorry that this point was written in the opposite direction. In the spirit of what you just said, we cannot have a specific discussion without looking at the details, and you pointed out that both are possible in the conclusion, so the conclusion has been rewritten in the opposite direction. Therefore, I would like to include the nuance of continuing while writing accurately.

Member : Thank you very much. The roles of natural persons are different between (1) and (2), so I would like you to describe them as you just pointed out.

Mr. Kozuka: Thank you very much, Mr. . Mr. Inadani is raising his hand, so please take care of him.

Inadani Member: Incentive, but since you nominated me, I will explain that story briefly again later. Before that, I would like to confirm a little. In your conversation with Mr. Takahashi earlier, is it okay to understand that the content of the report (draft) will not be revised in particular? I believe that the story you were concerned about earlier is basically reflected in the draft, so I do not think it is necessary to reflect or change it in particular. Is that okay?

Member Takahashi: It is at the end of page 27 of the Yes, I think that understanding is fine. I just expressed my impression.

Inadani Member: As Dr. Nakahara mentioned earlier, and on the premise that the Guidelines will be updated with better content as appropriate, I believe that the report (draft) this time is made on the basis of a balance between reducing the atrophy effect and making safe products at the same time by appropriately considering them in the interpretation of the reasons for liability as long as they are properly fulfilled. Therefore, in my opinion, it is better not to move them, and I have no objection to the point that they do not move.

In addition, in terms of incentives, to put it briefly, I believe that it will basically be possible by combining the two systems. One is that the sooner you cooperate, the lower the amount of monetary penalties such as surcharges will eventually be, and you will not have to worry about becoming the subject of an investigation, particularly a compulsory investigation. The other is that by preparing a whistleblower protection and incentive system, you will strongly recognize that if you leave bad things in the company, the company will suffer a larger economic loss, and as a result, you will be able to collect appropriate information while reducing the burden on the company and public entities related to investigations and investigations, so that appropriate procedures will proceed based on it. That is all.

Mr. Kozuka: Thank you very much, Mr. . Now, I would like to summarize your opinions, but before that, I would like to ask Mr. Hatano's opinion about the data collection in general, although you just announced the data collection at the beginning today.

Member of Hatano: Thank you for choosing . Overall, we would like to thank you very much for summarizing various opinions from four major perspectives, namely the Penal Code, the Civil Code, the Administrative Law, and accident investigation and analysis, and for reflecting them in the report (draft) in an exquisite balance. In particular, we would like to thank you for your consideration of the problems of the industry. At the same time, I am proud that we were able to have discussions that contributed to future consideration by having careful discussions as I mentioned earlier. There are no major points pointed out, but from the perspective of more effective rulemaking, we hope that more detailed and specific discussions will be held continuously, while taking into account the intentions of the business side, that is, the industry side. In terms of the relationship with infrastructure, I think there is still room for discussion on the demarcation of responsibility. Regarding the machine-readable version of the Road Traffic Law and the context of digitalization, which was talked about today, automated driving is a very effective means from the perspective of observing the rules. On the other hand, from the perspective of coexisting in a transportation society, I think there is still a certain Issue left in the fact that only automated driving and machines are provided with new rules. If human drivers cannot follow the strict rules due to the stricter traffic rules, there will be concerns about the double standardization of traffic rules. I would like to continue further discussions while carefully considering such matters. Technically, there are members who have expertise in various areas more than I do, so I would like to hear from you according to each theme and continue the discussion. Thank you for your time today.

Mr. Kozuka: Thank you very much, Mr. . I would like you to continue the dialogue with the business community in a careful manner. Next, Mr. Takahashi, please.

Member Takahashi: It is at the end of page 27 of the I am not saying that I want everything I said today to be reflected in the report, and I believe that the report (draft) is itself well done. One thing I would like to point out in advance is the relationship between medical care errors, which is probably a civil liability that will be raised in the future. When you get into a car accident, you may get a head injury, be transported to a medical institutions, and make a medical institutions mistake. If so, we will sue both the driver and medical institutions for joint tort. Such a thing often happened in the past, and the question of what the percentage of fault between medical institutions and the driver is became a problem, and finally, in 2001, the Supreme Court settled the matter. That is, in relation to the victims, both medical institutions and the driver will compensate the full amount. Of course, it is a joint and several liability, so if one party pays, the other party is discharged. But after that, medical institutions and the insurance company of the driver will make appropriate claims against each other. I think that when automated driving becomes, the same problem will be raised by medical institutions, so I pointed out that. That is all. medical care

Mr. Kozuka: Thank you very much, Mr. . I think that's all the opinions from the teachers. Regarding this report (draft), basically, we have received various compliments from the teachers, and this is due to the efforts of the Secretariat, but I think it is very good. Therefore, I think that the base can adopt this as the report of this sub-working group.
As one of the points you pointed out today, I think it would be better to add that, first of all, whatever organization is established for the investigation of the accident, it would be better to emphasize that it is different from the pursuit of responsibility. I think it would be better to emphasize this, including that there would be no misunderstanding if it were to go out to the general public who are not lawyers.
Second, Dr. Suda pointed out that there is a problem of who is responsible for updating safety standards, etc., so I think it might be better to touch on it a little. You may not be able to decide who is responsible, but it might be better to touch on it including the problem of who is responsible.
Also, machine-readable rules have come up in various contexts. In relation to criminal cases, for example, in the four quadrants on page 23 of the report (draft), you pointed out the relationship between the standards and guidelines, and Mr. Takahashi pointed out the relationship between Article 709 of the Civil Code and the Self-Compensation Act. At present, I think it is possible to narrow down the issue theoretically, but the civil court is making such an operational judgment, and in a sense, when it comes to the digital world, I think it is an issue that includes whether there are parts where what the judge was doing cannot necessarily be carried out. Machine-readable rules are written in the report (draft), but I think it is a very important issue, and I would like you to emphasize in a way that is easy to understand that this is something that must be addressed in the future, including those involved in practical operations.
On the other hand, there were various opinions at the time of today, but I don't think a consensus has been reached. One is how much compulsory authority should be given to the accident investigation agency. I received your opinion on that, but I don't think a conclusion has been reached at this point. Second, I don't think there are any objections to providing appropriate incentives to the people concerned in general, but I don't think we can make a final decision on, for example, how to evaluate the Deferred Prosecution Agreement System, because there seem to be differences in opinions. In addition, I don't think a consensus has been reached on how to write about the remote monitor in footnote 53, which you pointed out earlier, so I don't think we have discussed it much, but I would like you to write it so that there is no misunderstanding because it is a matter of how the remote monitor will pay attention.

We have received some major problems. For example, Dr. Nishinari mentioned that special consideration may be necessary when a malicious attacker attacks. You also pointed out the use of location information by mobile operators. In a broad sense, it can be said that this is a problem of the use of information on the infrastructure side. On the other hand, there are guidelines for mobile phone location information in Ministry of Internal Affairs and Communications, so I think it is a problem that has a large impact.

In addition, how to provide appropriate incentives to the people concerned in the future and whether it is in line with the idea of agile governance that Professor Inadani emphasizes are very big points to be discussed, but I don't think it is possible to launch it head-on in this report. Therefore, I would like the Secretariat to consider, for example, listing the parts that can be listed as future Issue. If listing them will be a problem in the future, I would rather not mention them, and I would like to keep them in the record as discussed in the minutes.

What I understand is roughly what it is now, and I would like to include these minor additions and revisions in this report (draft) and finalize it as a report from the sub-working group. Regarding the adjustment, I would like to finalize it after consulting with the Secretariat, so please take care of it.

There will be a show of hands from Dr. Takahashi, so please do so.

Member Takahashi: It is at the end of page 27 of the Report (draft), but is the sub-working group still continuing?

Mr. Kozuka: Thank you very much, Mr. Regarding the future, the Secretariat will explain, but as Mr. Goto pointed out, I believe that follow-up is necessary in some form.

This report will basically be compiled by me and the Secretariat. There may be another confirmation from the Secretariat, but is it okay with everyone?

Thank you very much. Then, I will do so. Then, as the time to finish is approaching, I would like to receive a summary from Deputy Director General Hasui as usual. Thank you very much.

Mr. Hasui: Once again, I am Hasui from . Thank you very much for your intensive and very clear discussions on the six occasions since last December until today. At the beginning, the agenda was difficult and the future was uncertain, but we received sincere discussions based on the high insights of the members, input based on the actual situation at the development site, and valuable opinions from the association of victims of the accident. Mr. Kozuka, Chief of Staff, summarized the important issues and gave instructions on the direction of consideration. As a result of your careful discussions, I believe it was very significant that engineers and legal evaluators came together to discuss the four fields of criminal, civil, administrative, and accident investigation. As a result, important issues essential for the future development of automated driving were organized, and I believe you made very valuable recommendations. I would like to express my deep gratitude once again. Digital Agency

Originally, at the Digital Administrative and Fiscal Reform Conference, you pointed out the relationship of responsibility in the automated driving incident, and we decided to discuss it. Therefore, we plan to report the content compiled today to the conference based on the final adjustment. In addition, in response to this recommendation, the ministries and agencies concerned will review the safety standards and guidelines and deepen specific discussions on systems design related to the investigation of the cause of the incident. This summary is located in mobility, which is coordinated by the Mobility Roadmap Working Group of the main body to which this sub-working group belongs. In Mobility Roadmap, a final decision is scheduled to be made in Council for the Promotion of a Digital Society, and based on this roadmap, the Government of Japan will roll the status of its efforts going forward. We also plan to have relevant ministries and agencies report on the results of their consideration of the content compiled by the sub-working group. Digital Agency will take the lead in cooperation among relevant ministries and agencies, and the Government as a whole will firmly respond. With the compilation of this report, the work of the Sub-Working Group will be completed. We would like to ask for further cooperation from the members of the Sub-Working Group, the Ai No Kai, the people involved in the automobile industry, and the people involved in the software and systems industry toward the acceleration of social implementation in automated driving. .

Mr. Kozuka: Thank you very much, Mr. . I would like to thank you again for your great cooperation in the six meetings and for making it this far. I would like to thank the Secretariat. As Mr. Ochiai said today, what we are struggling with is that we want to reduce traffic accidents in Japanese society as much as possible through automated driving technology. I would like you to clearly communicate the significance of this.

Lastly, on a business matter, as usual, the materials for today's meeting will be published on the Digital Agency website at a later date. The minutes will be published on the website after being confirmed by the members. Thank you. I heard that the Secretariat will make a notification here, so Councilor Kodama, please.

Councilor Kodama: Thank you very much, Mr. . As you said earlier, I understand that future reports will be left to the discretion of the Chairman. Is that okay? It seems that you have no objections, so I would like to proceed with the matter as if it were up to the Chairman. Regarding the draft report that I sat down with you today, I would like to share only the final revised version of the report without disclosing it, so I would like to ask for your understanding. That's all from here.

Mr. Kozuka: Thank you very much, Mr. .

Now, the 6th age of AI Sub-Working Group on Social Rules for automated driving vehicles has come to an end. Thank you very much.