Mobility Working Group (16th)

Overview

Date and

Wednesday, May 20, 2026, from 14:00 to 16:00

US>

On-line

Agenda

1. Opening
2. Agenda
- (1) Reporting and Dissemination of "Level 4 Model Collection on Ensuring Reliability of Communication Systems for Promoting Social Implementation of Autonomous Mobile Services"
  - (Ministry of Internal Affairs and Communications Bureau of Information and Communications)
- (2) Status of Initiatives for Priority Measures to Implement Social Rules for Autonomous Vehicles
- (3) Summary of Mobility Roadmap 2026
3. Exchange of Views
4. Closing

Material

Minutes

Proceedings of the 16th Mobility Working Group

Director for Policy Planning Fujisawa (Secretariat): , The 16th meeting of the Mobility Working Group will be held from now on. Thank you for taking time out of your busy schedule to attend the meeting today. I am Fujisawa from the secretariat, and I will serve as the moderator today. Nice to meet you. Today's meeting will be held in a hybrid format of online and online. All members are requested to keep the camera on during the meeting and to unmute the microphone when speaking. Please mute your microphone when others are speaking. Audience members are requested to turn off both the camera and microphone. We will record the meeting proceedings to share them within the secretariat and to prepare the meeting proceedings. We appreciate your understanding. Next, I would like to confirm the materials. As stated in the agenda that was sent in advance, there are the agenda, materials 1 to 5, reference materials, and the list of attendees. Also, today, regarding the recommendations of the project team on the future vision of autonomous driving hosted by State Minister Imaeda, separate materials from the meeting materials have been distributed. These materials will not be disclosed to the public, so please be careful when handling them. If there is anything missing, please contact Chat function or the secretariat by e-mail. Regarding the introduction of today's attendees, there are time constraints, so please check the list of attendees in front of you. Now, State Minister Imaeda is here today, so I would like to ask for his remarks before the opening of the working group. State Minister Imaeda, thank you.

Parliamentary Senior Vice-Minister Imaeda: Hello, everyone in . Thank you very much for taking the time out of your busy schedules to join us today. My name is IMAEDA Soichiro from State Minister for Digital Transformation. Today's meeting of the 16th Working Group on the Mobility Roadmap 2026 marks a major milestone in our deliberations thus far. I look forward to working with you all.
In addition, prior to the discussion, the project team for the future vision of autonomous driving, which I hosted, had various discussions in the past meetings and summarized them. The summary has been distributed to you.
I am truly grateful for the various discussions that have been held by people involved in mobility working up to now, and I have accepted such feelings. At the same time, looking ahead to 2040, the government has set a target of 10,000 vehicles by 2030. In the future, autonomous driving will be implemented in society and used in general, and of course, it will be commercialized in various vehicles. While drawing a vision of what the future will look like, I have considered what measures are necessary in backcasting. I would be grateful if you could understand that this is based on your discussions up to now and looking further into the future.
Autonomous driving is expected to make a solid contribution to social issues such as labor shortages, traffic gaps, and the reduction of traffic accidents, and is extremely important from the perspective of maintaining the competitiveness of the Japanese automobile industry. After identifying 17 strategic areas, the Takaichi administration is working to formulate a public-private investment roadmap by selecting products and technologies that are considered to require priority support for necessary public-private investment, and autonomous driving is also a major theme. I believe it is necessary to create specific measures to promote autonomous driving while keeping in mind the future vision I mentioned earlier, and today's discussion will be toward the Mobility Roadmap 2026, and I would like to receive opinions that will contribute to that.
In addition, an important point in implementing autonomous driving in society is the ideal form of liability systems and social rules in the event of accidents. I would also like you to discuss the status of consideration by the relevant ministries and agencies. Thank you very much. The roles that mobility working should play are becoming even more important, so we would like to place importance on this and firmly advance it, so we would like to hear your frank opinions. That's all. Thank you very much.

Director for Policy Planning Fujisawa (Secretariat): . Now, Chief Uno will open the Working Group. Chief Uno, please.

Chief Uno: Working Group. Today, first of all, I am from Ministry of Internal Affairs and Communications. I would like you to report on a collection of models related to ensuring the reliability of communication systems to promote the social implementation of Level 4 autonomous driving mobile services, and I would like you to give us your opinions. Next, I would like the Secretariat to explain the status of efforts on priority policies for the implementation of social rules related to autonomous vehicles and the Mobility Roadmap 2026 (draft), and I would like you to discuss them. Based on the opinions we have received from you in the working group so far, we have summarized the analysis of the current situation and recognition of challenges related to autonomous driving inside and outside Japan in this Mobility Roadmap 2026 (draft), and after summarizing the progress of each priority measure positioned in the Mobility Roadmap 2025, we have compiled measures to be taken from this fiscal year. Based on the opinions we have received from you today, we would like to aim for a ministerial-level decision on Roadmap 2026 at Meeting for the Promotion of a Digital Society, which is scheduled to be held in the future. I would like you to give us your frank opinions today as well, and I would like the relevant ministries and agencies participating today to firmly work on the implementation of the roadmap based on the opinions of the members. If there is anything that needs to be explained at this meeting, I would like you to speak from each ministry and agency. Thank you.

Director for Policy Planning Fujisawa (Secretariat): . As the proceedings dictate, we have three items on our agenda today. We will exchange views collectively after the conclusion of the three proceedings.
As the first explanation, Mr. Takada Kacho of the Regional Communications Promotion Division, Administrative Bureau of Information and Communications, Ministry of Internal Affairs and Communications, will explain about the report and dissemination of the "Model Collection on Ensuring Reliability of Communication Systems to Promote Social Implementation of Level 4 Autonomous Mobile Services." Thank you.

Takada Kacho (Ministry of Internal Affairs and Communications): Thank you, (Explanation of the report and dissemination of Document 2, "Level 4 Model Collection on Ensuring Reliability of Communication Systems to Promote Social Implementation of Autonomous Driving Mobile Services")

Director for Policy Planning Fujisawa (Secretariat): Thank you. Questions and answers will be summarized later. Next, Document 3 "Status of Efforts on Priority Measures for Implementation of Social Rules for Autonomous Vehicles" will be explained by Director Digital Agency. Thank you.

Suga Director (Digital Agency): (Document 3: Explanation of the status of initiatives for priority measures to implement social rules for autonomous vehicles)

Director for Policy Planning Fujisawa (Secretariat): , thank you for your explanation. Now, the third agenda item. Material 4: Outline of Mobility Roadmap 2026 (Draft) Material 5: Mobility Roadmap 2026 (Draft) Deputy Director-General Digital Agency will explain. Thank you in advance.

Deputy Director-General Okada (Digital Agency): (Document 4: Outline of Mobility Roadmap 2026 (Draft), Document 5: Explanation on Mobility Roadmap 2026 (Draft))

Director for Policy Planning Fujisawa (Secretariat): Thank you for your explanation. Then, I would like to receive opinions and questions from the members. Since there are time constraints, I would like you to speak for about three minutes per person. In addition, when you speak, please press the chat raise button or send a chat function. Those in the room, please raise your hands as appropriate. Mr. Yamamoto, your hand has been raised. Thank you for your cooperation.

Mr. Yamamoto: I am Mr. Yamamoto, Chairman of ITS Japan. Thank you very much for your overall explanation. I would also like to make three comments and requests.
First of all, on behalf of ITS Japan, I would like to thank you again for your attention to ITS infrastructure. In particular, the explanation by Deputy Director-General Okada on the Mobility Roadmap, which you explained at the end, was explained in a comprehensive manner, and I deepened my understanding. Among them, I would like to give one comment on the way of expressing or the content of ITS infrastructure. This time, with regard to communication, partly due to the effort with Mr. Ministry of Internal Affairs and Communications, it has attracted a lot of attention. I think communication is also necessary for remote monitoring. However, for example, we need to carefully determine what happens when we run automatic driving in a city or at an intersection. Looking at future technological progress, something like hazard prediction will become more realistic, including the progress of AI. We expect and assume that an information-processing infrastructure in which vehicles, people, and infrastructure cooperate will be one of the ITS infrastructures in the future. I would be grateful if you would consider information-processing infrastructure in addition to communication.
The second issue is related to communication, which was explained by Director Suga, but communication will always fail. Unfortunately, there are no communication environments that are free of failures. Each telecommunications carrier is doing everything they can to eliminate it, but it will definitely happen. Even if a failure should occur, I think it is necessary to prepare a minimum emergency communication line. This is an emergency situation. However, the other day, there was an incident in Wuhan where 100 Baidu cars stopped. That is an example of an automatic driving car stopping due to a communication outage. Therefore, as long as automatic driving is operated, communication is a prerequisite, and I think it will be one of the rules for rulemaking in the event of a failure, so I think it would be good to add this.
Finally, it is necessary to make fair decisions when accidents occur, or to clarify where the responsibility lies. In such cases, I believe that road infrastructure that constantly monitors or records the locations where self-driving cars run is necessary, so this is also related to rule-making, and I was concerned about it, so I would like to comment. That's all from me.

Director for Policy Planning Fujisawa (Secretariat): . Next, Mr. Koda, I would like to ask for your kind attention.

Member Koda: Thank you for your explanation. I would like to ask you a few questions. When you selected the 12 areas, I thought it would be easy to understand the background of the function area in Chitose. However, assuming implementation, I think it would be better to consider demonstration in areas where public transportation has not snowed at all due to a lack of population or economic inviability, or conversely, areas where traffic accidents frequently occur in urban areas. I also thought it would be better to have a model collection of single-track roads in rural areas or passing roads in rural areas when considering implementation, not just the demonstration of the current main roads. I would like to know if there is a prospect for that.
In another question, I would like to know in which areas the number of 10,000 cars by 2030 is expected to be achieved, if you have any. Looking at the current situation of local public transportation, not only elderly people, but also those who have abandoned their cars or cannot drive are in a situation where they cannot live their daily lives. If we aim to reduce the number of 10,000 cars by 2030 by returning licenses or reducing accidents for elderly people, we need to raise the awareness of the target people. Otherwise, where will the unnecessary demonstration be left as it is and we will enter the implementation phase? There are over 80 million cars in Japan now, and when I look at the population distribution and consider the elderly people, I would like to know if there is a vision of 10,000 cars by 2030, not to say 80 million cars, to solve which issues, and how much more I would like to spread.
Finally, I would like to ask about Suga Director. I think guidelines have been created, but since the service launch, regarding programming updates, I wonder if OTA is a mandatory introduction, and of course, if the car gets old, the situation where it cannot be installed even if the programming is updated will naturally occur. If there are measures to prevent accidents from occurring with old self-driving cars, I would like you to share the handling of old cars, who is responsible for cars whose programming can no longer be updated, and how to take responsibility. That's all.

Director for Policy Planning Fujisawa (Secretariat): Thank you very much. Mr. Hidaka, please start.

Prof. Hidaka ( ): Four points in short. The Ministry of Internal Affairs and Communications Model Collection is very easy to understand, and I think it is good to systematize the verification test. Probably, network slicing involving multiple layers will be done, and it will be enhanced. Then, it will be difficult to understand what will happen when there is a combination of effects such as mountains if it is isolated in the model collection. Therefore, I think it will be possible to organize what is connected when so-called systemic accidents occur in a form other than a tree structure. I think it would be good to include a paper referring to Professor Takeyama of Keio University's expression method when multiple elements of systemic design are involved.
The second point, which is a sub-working, I think that it is sufficient in the sense that it is made in Japan and used in Japan, but when something made overseas is used in the future, including equipment and vehicles, even if it is not intentional, it is assumed that it will be difficult to respond even if a substantial claim is made, whether the government will bear the compensation for the person who was run over in an accident, or whether there will be a move that the government will need to compensate to some extent, or whether there will be compensation in the event of an accident when no one in Japan is responsible for communication interference, etc. I think that there is a way to make accidents unique to automatic driving. This is a comment.
The third point is about the roadmap. As for the recent subsidized projects and autonomous driving demonstrations, there have been severe revisions by local governments, saying that if they are subsidized, they must introduce it. There are also some areas where local governments are reluctant to introduce it if they cannot. However, there are still R & D elements, including details such as the safety of autonomous driving and the opening and closing of bus doors. If we provide subsidies to support the introduction of capital investments, we must introduce them or we will have to deal with refunds. However, if we produce results in R & D and aim for something like the Ministry of Internal Affairs and Communications model I mentioned earlier, we should separate those with research elements from those to be introduced and continue R & D and verification test where more investment is needed.
Lastly, regarding the transportation trading companies, as Okada-san mentioned earlier, it is difficult to industrialize only with autonomous driving, and in particular, the transportation infrastructure, which can be said to be a Japanese national crisis, is collapsing. We are trying to do something quite difficult, so I think we included the transportation trading companies in order to include ideas such as joint procurement within the local transportation trading company function or the entire autonomous driving industry, or digital public goods for the Digital Agency. I start with the fact that the transportation trading companies are doing good things with mobility in the local area, but at the beginning, the first idea of the transportation trading company is how it can contribute to productivity, ease of introduction, and the digital domain. I think the idea of the cooperation promotion organization for the revision of the Regional Traffic Law is similar, and I would like you to include a word in the roadmap about the knowledge of the introduction part in the form of linking automated driving, which is an idea that came up during Digital Agency's working, and where the transportation trading companies will contribute.

Director for Policy Planning Fujisawa (Secretariat): Thank you very much. Then Mr. Nemoto, please start.

Basic Member: Please allow me to make two comments.
First of all, I get the impression that the terms "autonomous driving service business model" and "transportation trading company function" are explained independently, but I think we need to sort out the relationship. When we think about business models for buses and taxis in underpopulated areas, I think we have learned that it is desirable to have an entity that plays the role of a transportation trading company function. On the other hand, for autonomous taxis in urban areas, an Uber-type business model has already been established as a de facto in the U.S. and Chine, and for autonomous trucks on expressways, it has started in the U.S. and Chine, but I don't know what kind of business model is suitable yet. There are various ideas such as the need for a switching base of a public nature or the creation of an organization that centrally monitors remotely and cooperates well with NEXCO's control center. However, I think it is necessary to discuss what kind of business model should be used to establish the project in the future. What I wanted to say was that a small part of the autonomous driving service business model is transportation trading company function.
Another point is that when explaining the consolidation base for logistics-related measures, the term "for switching and transloading" is used, but I personally think that the term "switching" is more appropriate. Because the term "switching" means that a single-car is assumed as an autonomous vehicle, and when the single-car switches from manual to automatic, the driver gets off. However, as in Europe, the United States and China, they think it is better to transport long distances by semi-trailer, in that case, it will be the replacement of the head and the trailer. I think that the term "switching" from manual to automatic driving, which can include both of them, is better than "switching". I think that such a term is being used more often, so how about doing so?

Director for Policy Planning Fujisawa (Secretariat): Thank you very much. Mr. Kawabata, please start.

Prof. Kawabata: I would like to ask a few questions.
I think you have illustrated various patterns of responses to automated driving when an accident occurs, but I do not understand the criminal punishment in the case of Level 5 remote control, and in the current system, it seems to be appropriate for criminal punishment, but when I was in charge of, for example, 10 or 20 vehicles remotely, and a serious accident occurred, I felt that it was complicated about who was at fault and whether it was the system's responsibility. I feel that such discussions were held in some committee, and I would like to know what the final form is at present.
In addition, as I think other members also pointed out, the weight of the private sector is high in urban areas, so the commercialization of autonomous driving will be possible, and it is also global, but the biggest challenge is still in rural areas, and in some cases, even if it is not advanced E2E, it may be enough to drive slowly with image recognition in rural areas. There has not been much discussion on such technology selection, so if that cost is not spent in rural areas, when E2E's advanced autonomous driving is introduced, it will not be monetized and the business model will stop, which is close to stopping when verification test ends. I think it would be better to touch on how to select appropriate technology for that area. If you have any ideas, I would like you to talk about that as well. It is directly linked to the commercial vehicle model, and as I mentioned earlier about trucks and logistics, when large buses, on-demand transportation, taxis, etc. are introduced in rural areas, the ratio of public and private sectors inevitably comes out, which directly leads to costs, and how much advanced technology can be suppressed, I think this discussion is necessary.
One more thing, numerical targets are important and I think it is better to be ambitious, but when you show what kind of startups there are in E2E, even though there are multiple startups and the scale is getting bigger to some extent, when you set a target of tens of thousands of units, I think there is a desire for startups to develop technology. I think there is a bit of a lag against China and the United States, but I think startups can be exhausted by the number of units target despite wanting to promote technology development, so I wonder how to think about this balance.
Finally, I have one more question. Digital technologies and public frameworks have been organized, but as social infrastructures, I do not think I can yet talk about organizing physical issues, such as, for example, you cannot drive because there are steps, or you want to install bus stops but cannot do so in front of the station. However, when it is commercialized after 2030, there will be cases where people will not use it because it is not easy to use, or they will not use it because there are steps, or they will not use it because they cannot build bus stops in front of the station. It takes more time and money than digital to develop a physical platform, so if you have any thoughts on this, please let me know. If you have not reached that point yet, I thought it would be better to mention it somewhere.

Director for Policy Planning Fujisawa (Secretariat): Thank you very much. Mr. Akimoto, please start.

Deputy Member Akimoto: This is Akimoto of the Japanese Unmanned Vehicle Operation Management Consortium. Thank you for your explanation. I would be grateful if you could create this roadmap with a vision of eliminating traffic accidents, including preventing accidents for elderly drivers, by introducing and implementing autonomous driving in society, and create a society in which there are no more accidents. It is thought that the elimination of accidents will reduce the insurance premium rate for compulsory automobile liability insurance and voluntary insurance. I would like to hear your thoughts on this.
I have two more questions. As you explained in the roadmap, Japanese companies are lagging behind in E2E AI development and AI introduction. Against this backdrop, Chinese companies are accelerating the development of various technologies in the context of competitions, such as a marathon for bipedal robots. In Japan, for example, we are introducing self-driving vehicles from overseas, or creating a test site for self-driving vehicles for companies that are developing various technologies in Japan, and competing there, comparing the level, while looking at various ideas. I think Japanese companies are good at building reliability, so while looking at various issues, I think it is Japanese companies' strength to build reliability, including reverse engineering in the past. I think it is also important to compare the level of what each company is doing at competitions and test sites.
With this in mind, if there are categories such as technological development in the future roadmap, I think they will be included and contribute to the progress of technology, and the introduction of generative AI will also be accelerated.

Director for Policy Planning Fujisawa (Secretariat): Thank you very much. Ms. Wakana, please start.

Member WAKANA: I'm Wakana from the Iwate Regional Development Support Center, . I have one request, and it's about Transportation Trading Company function. There has been a lot of discussion, so I have one request. On page 32 of Document 5, it says that the case studies will be revised, and I wonder if you could add to it, including the discussion you just had. As for the reason, as for the way autonomous driving will be implemented in the field, as other committee members have mentioned, in urban areas, it will include a form similar to Uber. However, in areas with a small population in rural areas, I think it was pointed out in past meetings that it would not be economically viable. At that time, I think it would be more realistic to include Transportation Trading Company function as presented here on the premise of sharing vehicles, and I think it is Transportation Trading Company function for that purpose. As explained earlier, the main thing is that the Ministry of Land, Infrastructure, Transport and Tourism will revise the Regional Traffic Act, set up collaboration promotion organizations, and although it may be similar, we will collaborate on the supply system. Transportation Trading Company function, as has been discussed for a long time, I think there will be needs that will be consolidated and shared, and I would like this to be promoted by Digital Agency.
As I commented in the previous discussion, as you just pointed out, I would like you to go into the linkage between transportation trading company function and autonomous driving. As I said in the previous discussion, I believe that the digital infrastructure associated with the introduction of autonomous driving will contribute to solving social issues in some ways, but the example you presented is the introduction of transportation trading company function, and page 32 of Document 5 also describes the horizontal development of transportation trading company function, but I think that's putting the cart before the horse. Automatic driving and the digital infrastructure associated with it will contribute to society in the form of transportation trading company function, and it will lead to solving social issues. If you don't add a little more analysis of the examples in the current collection of examples, and if you don't add another example of what kind of transportation trading company function assembly will lead to solving issues, I think it will be a waste of the results of the survey you have done. I hope that you will add a little more to this in the required revisions.

Director for Policy Planning Fujisawa (Secretariat): Thank you very much. Mr. Suda, please start.

Member: Tokyo University of Technology, My name is Suda, and I would like to make three comments and ask questions.
First, the original purpose of autonomous driving is to ensure safety. However, I feel that it is a little biased toward the recent confrontation between E2E and rule-based systems. E2E is certainly an attractive and practical story, but the purpose of autonomous driving is originally safety. Humans make errors, so the story was about keeping safety by following traffic rules in an automated and rule-based manner. However, in the direction of imitating humans with E2E, it goes back to the past. I thought it would be good to have a proper discussion on that and describe how E2E should be used, not as a conflicting concept between E2E and rule-based systems, but as a hybrid that needs to take the advantages of being able to use it well.
In addition, the target of 10,000 vehicles is vague, and it is unclear what the breakdown of 10,000 vehicles is. In implementing autonomous driving, there are differences in algorithms, such as whether only vehicles should be autonomous or whether infrastructure should be coordinated. As for the level of autonomous driving, is 10,000 vehicles only level 4 or level 2 + +? The target has been buses, trucks, and service cars until now, but robotaxis have been established overseas, and I thought it would be good to clarify the story around this, such as whether to aim to be an owner car in the end.
There are still areas where the use of signal information has not been realized at level 4, and I think there are great expectations for this.
Finally, I hope that the Act on Revitalization and Rehabilitation of Local Public Transportation Systems will be amended. I was at the meeting of the Council for Transport Policy, and the university is working with the Hachioji City Hall, so I would like to see the contents of this as much as possible.

Director for Policy Planning Fujisawa (Secretariat): Thank you very much. Mr. Ishida, please start.

Ishida: In regard to the writing of the Roadmap, there is a recognition that the urgency of time is becoming more and more severe year by year. In terms of industry, the gap between Japan and China and the United States is widening year by year, so the strategy is good, but I think it would be good if the points of contact at the lower levels of the roadmap, such as tactics and logistics, could be written a little more.
Related to this, mobility from the community side is also in a difficult situation. The well-being of elderly people who have returned their licenses will decline, and the community will lose vitality, or there will be no public transportation. I think safety is the most important thing, but recently traffic accidents involving pedestrians, bicycles, and road bikes, called VRU (Vulnerable Road Users), have turned to an increase. I think it is important to take this into consideration and include the perspective of VRU as well as automobiles.
If we race against time, for example, there is an advanced driving support called L2 + +, so I think it is important to figure out how to set up a platform as a public good to support the introduction. There are examples of how we have worked as a national policy, such as the process of introducing computers and expensive medical equipment, so I think that is possible.
Or, I would like to digress, if we consider automated driving as a mobility service instead of lumping them all together, the operation and quality of mobility services for transporting people and mobility services for transporting goods are different, so I think it is also important to understand this.

Director for Policy Planning Fujisawa (Secretariat): Thank you very much. Mr. Mori, please start.

Member MORI: , I would like to add a few comments based on the points made by each committee member so far. First of all, regarding function, a transportation trading company, the progress of autonomous driving and the support of public transportation in rural areas are not one on-one responses, but roles that contribute greatly to each other, so I would like them to naturally proceed. It is OK not only to create a collection of model cases, but I would also like you to consider in which fields the platform will be publicly supported. As expected, each local area supports public transportation or taxis, and the environment is quite severe, so I would like you to consider to what extent public support will be provided.
In addition, trucks may have a slightly different meaning from transportation trading company function, but regarding truck collection and delivery, I think it would be good to have some kind of system in the future. I would also like you to consider giving transportation trading company function in a form different from the current flow.
Lastly, in order to further advance the accident investigation, I think it is necessary to publish and share CAN data, which I have been talking about for some time. I think it is being discussed at the United Nations and I believe it is also being discussed at WP29, but when self-driving cars run in Japan, Tesla's cars will be fully compensated, but Japanese cars in Japan will not be compensated, so Tesla is disclosing data as an international standard. This is supposed to be done globally as well, but if only Japanese cars cannot disclose data, the accident investigation will not advance for Japanese cars, and I think Tesla would be better. If that happens, it will conflict with the goal of increasing the number of 10,000 and domestic cars I mentioned earlier, so I would like you to lay the groundwork and level the ground. As I have mentioned several times, it is advancing globally in the form of Uber in major cities. I would like each ministry to discuss to what extent public support can be provided to various vehicles such as buses in rural areas and trucks and passenger cars at the national level, and I would like you to proceed with the discussion from the roadmap to the next step.

Director for Policy Planning Fujisawa (Secretariat): Thank you very much. Then, Mr. Muramatsu, please start.

Member Muramatsu: I would like to tell you two things.
The first point is about communication. In the field of robotics, which I am in charge of, there are many robots that move indoors, so many studies are being conducted on how robots will move on the premise that communication is not connected. As for CPUs, there are many robots in the world that have solid CPUs on the edge side and on the robot side, and can move properly even if communication is lost. Robots are small, and there is a premise that the risk of injury to people is quite low, so this is the trend. Whether to move the function of CPUs to the cloud side or to the edge side is a matter of discretion, and I understand that there is no right answer, so I think it is necessary to repeat trial and error, taking into account the specifications and operation of autonomous vehicles. In specific terms, as another committee member mentioned earlier, I think it is important to define the degree of autonomous driving according to the circumstances, including communication, vehicle specifications, and circumstances. Is it important to limit the speed in areas where communication is poor and often interrupted, or is it important to at least load CPUs so that the car can move to a corner even if communication is interrupted? I am repeating myself, but I think it is necessary to discuss what to do on the premise that communication is interrupted.
The second point is, regarding the business model, I understand that it would be better for the future if we clarify the expenditure in more detail. When it comes to autonomous driving, communication with the cloud side inevitably occurs. When it comes to E2E, it is the same for communication and system maintenance on the cloud side. In the past, once the car was sold out, it was basically a maintenance business for another company, but I understand that it will be inevitable that a monthly usage fee for maintenance will be charged. Of course, I agree with strengthening function, a transportation trading company, and expanding the range of positive revenue, but how much expenditure will there be after all? I understand that it will be useful to clarify the expenditure for the portion that has not occurred after car sales in the future, when we examine the business model in the future.

Director for Policy Planning Fujisawa (Secretariat): Thank you very much. Mr. Okamoto, please start.

Okamoto, Member: I am Okamoto of TEPCO Holdings. Autonomous driving is expected to become a social infrastructure, and consideration for safety is increasingly important. Based on this premise, I would like to convey three points from the perspective of an infrastructure operator.
My first question is: As a technology, E2E is sophisticated, but I understand it. Recently, I also understand that doing it with VLA allows very flexible and sophisticated movements. On the other hand, I feel that it is a risk to steer everything to E2E. Considering that we cannot avoid hallucination even if we usually use generative AI, VLA will be the same, so I think it is necessary to set guardrails with logic and rules in real time. I think we have no choice but to consider combining it with logic.
The second point is, as Yamamoto-san mentioned earlier, I think we need to consider communications and information processing in an integrated manner. For example, information may be generated in the car, including sensors, or it may be sensed by the infrastructure side, but it will eventually be processed somewhere and lead to the next action of the car, so I think we need to create a service with no overall delay and a deterministic delay type. If we don't receive an answer by a certain time, I think we need to instantly move to a backup means. I don't think this is a best-effort, so if communication and information processing are not integrated and supported by a deterministic delay type service, the risk is too big, and I think we need to consider redundancy in that.
The third point is how to make good use of the existing infrastructure. For example, optical fibers are the dark fiber of communication carriers, the dark fiber of electric power companies, and the dark fiber of railway companies, so I think these can be used for communication and information processing infrastructure. In addition, utility poles, electric wires, and optical fibers are connected to substations, and by placing edge servers at substations, ITS can be provided by smart poles, and similar things can be processed at substations at utility poles. The optical fibers, electric wires, and poles there can be used as they are, so it is important to make the most of them. New investment would be huge, so I think how to use the existing infrastructure is required. We are proposing this in the form of Watt-bit linkage, and I hope you will consider linking Watt-bit linkage and autonomous driving.

Director for Policy Planning Fujisawa (Secretariat): Thank you very much. Mr. Hatano, please start.

Member Hatano: , I would like to express my gratitude to the secretariat and all the people concerned for their extensive discussions and for the hard work they put together in a short period of time by newly adding an autonomous driving section on AI.
However, looking back at the 2025 roadmap, self-driving technology has moved from the demonstration stage to the implementation stage. Therefore, based on the idea that it is important to ensure profitability in the future, I think we have exchanged opinions with the members of this year's committee on the consideration of making various mobility sustainable, including function, a transportation trading company in 2026.
I would like to reconfirm the position of the fact that autonomous driving technology has once again become the center of discussion and has been included in the roadmap from the perspective of E2E AI. In other words, I think it is a very important perspective as to what approach will be taken, whether it is necessary to demonstrate the technology again, or whether new technology should be considered in parallel with the project at the implementation level rather than demonstration. In particular, the technology aspect has been discussed in the Mobility Roadmap five times this time, but there has been no discussion of technology so far, so I think it is important to clearly organize the new addition and reflect it in the roadmap. This time, especially for E2E AI, the term L2 + + has been used to argue that there are challenges in spreading and expanding it, but unfortunately, the term L2 + + is not used internationally and is a term unique to Japan, so it is necessary to confirm again that there is no gap in recognition of the use of this term internationally.
In addition, L2 + + is only a driver assistance technology, and in the end, it is only a technology that the driver operates with safety responsibility. On the other hand, if we consider that what we must discuss is an autonomous driving technology that is L4 or higher that does not require a driver to maintain and expand traffic environments including local areas, whether the technology gap between L2 and L4 can be handled consistently with the term E2E in this discussion is a big technical challenge. We felt that it is important to clearly express and confirm in the roadmap, including in-depth discussions on why autonomous driving is not able to achieve sustainable penetration and expansion and why it is not commercially viable, rather than simply creating the context of a comprehensive technical solution with E2E AI, so I would like to ask for continued in-depth consideration.

Director for Policy Planning Fujisawa (Secretariat): Thank you very much. Mr. Hanami, please start.

Deputy member of Hanami: I am attending on behalf of Mr. Saito from IPA. I would like to comment on a few technical aspects. Assuming that a social function based on autonomous driving will eventually become an infrastructure, it is likely that not one system will be established in a country, but similar systems will be established in each Architecture and various units. To do so, it is necessary to establish a design with a solid focus on standardization from the beginning. In our organization, we use the expression "align", but for example, between vehicles and information infrastructure, between road surface infrastructure and information infrastructure, between vehicular information infrastructure and medical information infrastructure, and so on. I think we should also design with a solid focus on standardization for interfaces with various other fields related to the transportation trading company function. By aligning the module unit and model, we can respond to changes in hardware when the system changes generations after five or ten years. I think we need to do sustainable design from the beginning, assuming that it will eventually become a social infrastructure with a long span.
In addition, function, which is related to safety, is very important, so it may be necessary to introduce mechanisms such as certification and accreditation as well as standardization. There may be differences in the way they are made, but I think it is also necessary to design a unified method and way of thinking when a failure occurs, such as fail-safe and fail-ize when something happens. That's all.

Director for Policy Planning Fujisawa (Secretariat): members. Now, I would like to respond to the opinions received in the order of presentation. First of all, I would like to respond to Ministry of Internal Affairs and Communications's explanation. Takada Kacho, please.

Takada Kacho (Ministry of Internal Affairs and Communications): Thank you, . Regarding the question of whether there will be a selection of areas in hilly and mountainous areas as pointed out by Koda members, please refer to page 3 of the material. In Misato-cho, Shimane, we plan to conduct a demonstration in a hilly and mountainous area. In addition, in Miyazu City, Kyoto, due to its historical townscape, we assume that autonomous driving will be conducted on fairly narrow roads. In addition, we plan to conduct verification in various urban and rural environments, such as Ishigaki City and Komae City, Tokyo Prefecture, which is a big city. We will continue to provide solid verification based on the perspectives received.
In addition, I think it is reasonable that Member Muramatsu and Member Okamoto pointed out, "Shouldn't we take a firm fail-safe in an environment where the connection is not available?" This may be a part that was not fully explained, but under the Road Traffic Act, continuous remote monitoring is mandatory for Level 4 automatic driving. We are conducting a technical demonstration of how to perform continuous remote monitoring as necessary, sufficient, and efficient as possible, and we are fully aware that it is naturally necessary to consider separately what measures should be taken when communication is interrupted. I apologize for causing a slight misunderstanding from the perspective of communication. That is all.

Director for Policy Planning Fujisawa (Secretariat): , thank you very much. Next, Mr. Director Suzuki, could you explain the status of the sub-working?

Director Suzuki (Cabinet Secretariat): This is Mr. Suzuki from Cabinet Secretariat, . I would like to give the answer on behalf of Mr. Suga.
First of all, Mr. Koda said that in the case of an old car that can no longer be updated, a certain period of time will be set, even after the standards have changed, but if it still cannot be updated, it will not meet the safety standards, so I hope you will understand that basically the car will not pass the vehicle inspection.
The second point is about the accident compensation that the Hidaka members mentioned. As you know, self-driving cars are not currently available for sale, but the idea is to compensate within the scope of the so-called operator's liability. Therefore, in the case of self-driving cars, it will be the owner, and in the case of buses, it will be the bus operator, but in the case of operator's liability, the victim will be compensated.
Next, regarding the accident during remote monitoring that Committee Member Kawabata mentioned, in principle, the automated driving system and the vehicle itself are responsible for its operation, so the first question is whether the operation was within the design domain when the accident occurred. If it was within the design domain, we will look at whether it met the safety standards. In addition, regarding whether the designated automated driving supervisor was fulfilling the required responsibility, I am aware that it has been arranged under the Road Traffic Act that the remote monitor will be held responsible if there is serious negligence. That is all.

Director for Policy Planning Fujisawa (Secretariat): . I would now like to ask Deputy Director-General Okada to explain the Mobility Roadmap.

Deputy Director-General Okada (Digital Agency): As you have made comments on a very wide range of topics, I may not be able to answer all of them, but I will talk about them.
We received comments on Transportation Trading Company from several members. You said that it should be described clearly from the background and purpose of launching Transportation Trading Company function at the time of the roadmap in 2025, so I would like to review the description again.
With regard to function, a transportation trading company, how can we introduce automated driving in places where transportation is inconvenient, especially in rural areas rather than urban areas? Since it is difficult to simply introduce high-end automated driving, I think they have delved into it thoroughly from a business aspect as well. I would like to consider function, a transportation trading company, including the relationship between automated driving and digital services.
In terms of technology, we have included a considerable amount of new information on E2E, including the compilation of growth strategies this time. However, we do not believe that E2E alone will solve the problem, and page 13 states that we will continue to work firmly on rule-based and modular AI in parallel. You mentioned that even if E2E is introduced, it will be safer to include rule-based elements such as ITS. The overall tone is that we will work properly in parallel with E2E, including such aspects. I would like you to recognize that we will not advance everything with E2E alone. In addition, when I talked with various companies, I heard that many of them did not think that E2E would spread so rapidly. Based on the progress of technology, I may have felt a little sudden, but I believe that Japanese companies cannot afford to lag behind in E2E, and we have to take various initiatives in parallel. I would like you to understand that we have included several measures related to AI.
I believe there was an explanation on the bill, a breakdown of the 10,000 vehicles, and maintenance of physical bus stops and steps, but if there is anything, please contact the Ministry of Land, Infrastructure, Transport and Tourism.

Director Takamoto (Ministry of Land, Infrastructure, Transport and Tourism): This is Takamoto of the Distribution and Motor Vehicles Bureau, Ministry of Land, Infrastructure, Transport and Tourism. I cannot give a clear answer because it is related to the bus division and passenger business, but first of all, I would like to talk about the case of 10,000 units. This figure was set as a major goal in the 3rd Basic Plan on Transport Policy by the Ministry of Land, Infrastructure, Transport and Tourism in January this year. The figure of 10,000 is a very ambitious figure, and at the stage where only about 10 units have yet been granted permission for specific automated operation of L4 at present, in terms of setting a figure that is a thousand times that, I would like you to first understand that it is a figure set as a major goal rather than a world of accumulation. It is a figure based on various factors such as the number of taxis, buses, and trucks, but it is one of the target figures.
As for the future, I would like to work hard over the next five years to get as close as possible to this fairly ambitious figure of 10,000, including awareness. Therefore, I would like to consider the future based on the situation. I think it is a world view of how to satisfy this figure of 10,000.
Regarding the discussion on insurance premiums from Dr. Akimoto, I heard that each business operator has started to communicate with insurance companies on an individual basis. While it will be commercialized more widely, I think there will be a commitment to the government, but I think that each company will first proceed with things.
As Mr. Hitoshi Suga, Director of the Ministry of Land, Infrastructure, Transport and Tourism mentioned earlier, we will establish a system to investigate the accidents from this autumn toward next year. At the same time, we would like to build a framework to collect data from the operators in the same way. That is all from the MLIT.

Deputy Director-General Okada (Digital Agency): Thank you very much. There are some points that I haven't commented on other than that, but I would like to carefully review all of the opinions received and link them to the revision of the roadmap, so please check the revised version again.

Director for Policy Planning Fujisawa (Secretariat): .

TOYAMA Deputy Director-General (Ministry of Land, Infrastructure, Transport and Tourism): I would like to speak about the topic of physical road space from a member of Kawabata . From the perspective of autonomous driving, we are organizing a manual for the development of road space that makes it easy for autonomous vehicles to run smoothly. On the other hand, regarding the development of terminals that are easy to use for public transportation from the perspective of users, I think it is limited to autonomous driving, but we will consider making it easy to use from the perspective of business operators through the development of regional transportation bases and local public transportation consultative bodies. That is all.

Director for Policy Planning Fujisawa (Secretariat): . Mr. MLIT also gave a comment on the topic of the revision of the Regional Traffic Act and the relationship between the cooperation promotion organizations and function, a traffic trading company. Could you give us any comments on this?

Manager, Hoshi (MLIT): In the recent revision of the Regional Traffic Law, we have conducted a stipulation on the comprehensive adjustment of supply and demand, which has been discussed by traffic trading companies in the past, and the function of the corporations responsible for it. In addition, we are promoting stipulation as a mechanism, including institutional collaboration with other fields. Our council is also discussing it, and it is not very efficient to discuss it repeatedly, so we are currently discussing the roadmap shown with Mr. Digital Agency. We will further scrutinize and respond based on what you have pointed out. Thank you very much.

Director for Policy Planning Fujisawa (Secretariat): , thank you very much. Are there any comments in particular from other ministries?
Is there anything in particular? Thank you very much. Thank you very much for the lively discussion. As this is the scheduled time, I would like to ask Chief Uno to give a summary of the working group's closing.

Chief Uno: Thank you very much for your valuable opinions over a long time today. The opinions received are at the stage of this Mobility Roadmap 2026 (Draft), so we will reflect them as much as possible, and then proceed with the process of making a decision in Meeting for the Promotion of a Digital Society. I would like the relevant ministries and agencies participating today to cooperate in revising the draft roadmap based on today's discussions. In addition, I would like them to sincerely work on each measure presented in the compiled roadmap. In any case, I would like to make every effort to realize the early social implementation of autonomous driving.
With this, I would like to end our discussion on the Mobility Roadmap 2026. I would like to express my sincere gratitude to everyone for their cooperation so far. After this, the working group will continue to consider the progress of the initiatives taken up in the Roadmap 2026 and new issues that have arisen as necessary. I would like to ask for your continued efforts. Thank you very much for today.

Director for Policy Planning Fujisawa (Secretariat): Thank you very much for all the advice you gave me today. We were able to hold the 16th meeting thanks to your cooperation. Once again, the secretariat would like to express its gratitude.
We will compile a final plan for the Mobility Roadmap based on the opinions received. If you have any additional opinions, please contact the secretariat by the end of this week.
With regard to the schedule going forward, we will make necessary revisions, etc. based on the opinions received today in preparation for the decision to be made in Meeting for the Promotion of a Digital Society, which will be held in the future. During this period, we ask for your full attention to the handling of materials.
This is the end of the 16th Mobility Working Group meeting today. Thank you very much for your time today.

Greater than or