In the lead up to the 2020 Tokyo Olympics, Japanese automakers have set ambitious goals with the country, seeking to show its technological prowess in the field of Autonomous Vehicles. Numerous Japanese automakers have announced plans to have autonomous vehicles on display during the 2020 Tokyo Olympics; however, the level of automation that may be available on public roads is less than clear, particularly given the laws and regulations that have been in place.
Recent rules by the Japanese National Police Agency (NPA) are, however, fostering greater innovation with guidelines being put in place to permit self-driving tests on public roads.
In this section, we consider (1) the Japanese autonomous vehicle market; (2) the current Japanese legal and regulatory landscape (and future reform, strategy and guidelines); (3) Product liability and insurance; and (4) Cybersecurity and privacy.
Various concepts with respect to automated driving under Japanese laws and regulations
In June 2014, the Japanese government established the “Public-Private ITS Initiative/Roadmaps” a policy paper considering the introduction of intelligent transport systems (ITS) into Japan. This policy paper has been revised four times since 2014 with an ever increasing focus on automated driving systems in the country.
The development of the Public-Private ITS Initiative/Roadmaps policies has allowed for greater ITS-related development and innovation by Japanese ministries, agencies, and the private sector; particularly in respect to the promotion of specific collaboration among related government ministries, and the encouragement of competition and collaboration among private companies.
Since 2014, various concepts of “driving” have been defined with the policy whereby “driving” is stated in terms of the level of the driver’s involvement. The latest Public-Private ITS Initiative/Roadmaps 2018 (Roadmaps 2018) published in June 2018 adopts the definitions described in SAE (Society of Automotive Engineers) International’s standard J3016 (revised as of September 2016) for definitions of automated driving levels.
The overview of the definitions of automated driving levels adopted by Roadmaps 2018 are stated as follows:1
Driver performs part or all of the dynamic driving task (DDT)
No Driving Automation – The driver performs the entire DDT, even when enhanced by active safety systems.
Driver Assistance – Sustained and ODD (operational design domain)125 – specific execution by a driving automation system of either the lateral or the longitudinal vehicle motion control subtask of the DDT (but not both simultaneously), with the expectation that the driver performs the remainder of the DDT.
Partial Driving Automation – The sustained and ODD-specific execution by a driving automation system of both the lateral and longitudinal vehicle motion control subtasks of the DDT, with the expectation that the driver completes the OEDR (object and event detection and response) subtask, and supervises the driving automation system.
Automated driving system (ADS) performs the entire DDT (while engaged)
Conditional Driving Automation – The sustained and ODD-specific performance by an ADS of the entire DDT, with the expectation that the DDT user is receptive to ADS-issued requests to intervene, as well as to DDT performance-relevant system failures in other vehicle systems, and will respond appropriately.
High Driving Automation – The sustained and ODD-specific performance by an ADS of the entire DDT without any expectation that a user will respond to a request to intervene.
Full Driving Automation – The sustained and unconditional (i.e. not ODD-specific) performance by an ADS of the entire DDT without any expectation that a user will respond to a request to intervene.
In summary of the above, Roadmaps 2018 defines automated driving systems at Level 3 and above as “Highly Automated Driving Systems,” and those at Levels 4 and 5 are collectively called “Fully Automated Driving Systems.”
Moreover, according to SAE International’s J3016 (2016), automated driving systems can be divided into those with a user (including those who are the equivalent of drivers) who is inside the vehicle, and those with a user outside the vehicle, who remotely monitors and operates the vehicle. The Roadmaps 2018 defines the latter (a driving automation system with a user outside the vehicle) as a “Remote Automated Driving System,” and transport services that use such Remote Automated Driving Systems are defined as “Unmanned Autonomous Driving Transport Services.”
Road Traffic Convention and Law
Japan is a signatory to the 1949 Geneva Convention on Road Traffic, but not the 1968 Vienna Convention on Road Traffic. The relevance of this fact is that:
- Article 8 of the 1949 Geneva Convention states that “every vehicle or combination of vehicles proceeding as a unit shall have a driver” (paragraph 1), and that “drivers shall at all times be able to control their vehicles” (paragraph 5); and
- Article 10 of the Geneva Convention states that “the driver of a vehicle shall at all times have its speed under control and shall drive in a reasonable and prudent manner.”
Given this language, the Road Traffic Act of Japan also assumes the “existence” of a driver, and stipulates that the driver of a vehicle must work the vehicle’s steering wheel, brakes, and other equipment in a consistent manner, and must drive at a speed and in a manner that poses no hazard to others in consideration of road conditions, traffic conditions, and the condition of the vehicle.
According to Roadmaps 2018, the actual operation of vehicles on public roads in Japan is allowed for autonomous vehicles of Level 2 and below without infringement of the laws or regulations of Japan, provided that there is a driver inside
the vehicle who must handle the steering wheel, brakes, and other equipment, but not for autonomous vehicles of Level 3 and above.
Road Transport Vehicle Act
The Road Transport Vehicle Act of Japan provides that vehicles must not be operated unless they conform to the safety standards in respect of various features of the vehicle (including steering and breaking equipment) issued by the Japanese ministry, the “Ministry of Land, Infrastructure, Transport and Tourism” (the MLIT).
In February 2017, in order to enable field operational tests of unmanned autonomous vehicles on public roads, the MLIT revised and relaxed the safety standards under the Road Transport Vehicle Act to allow vehicles with no steering wheel or accelerator pedal on the premise that alternative safety measures (such as limiting the driving speed, limiting the driving route, and equipping an emergency stop switch in the vehicle) are taken.
Regulation with respect to field operational testing and actual operation of automated driving systems on public roads
Field operational tests on public roads and actual operation of automated driving systems in Japan is being developed based on international discussions to ensure consistency between global automated driving systems and the Geneva Convention on Road Traffic.
In May 2016, the National Police Agency of Japan (the NPA) announced the “Guidelines for Field Operational Tests of Automated Driving Systems on Public Roads,” which clarified that field operational tests on public roads regardless of the level of automation (i.e. Level 1 – Level 5) are allowed, without prior arrangement with or permission from the police, if:
- the vehicles operate in compliance with related laws and regulations, including the Road Traffic Act, and
- there is a driver in the driver’s seat who ensures that emergency situations can be handled.
Under these Guidelines, the person who assumes the role of the driver is required to:
- at all times be seated in the driver’s seat of the vehicle; and
- monitor the surrounding traffic, as well as the vehicle’s condition; and
- in the event of an emergency, operate the vehicle as necessary to ensure safety, and thus prevent damage to others.
Recent changes to field operational testing laws
The international discussions at the Global Forum for Road Traffic Safety (WP1) of the U.N. Economic Commission for Europe (the UNECE) confirmed in 2016 that there was no need for amendments to the 1949 Geneva Convention on Road Traffic, for foreseeable types of experiments (i.e. where there is someone ready and able to take control of the experimental vehicle, this person may or may not be inside the vehicle).
Based on such international discussions, Japan sought to develop institutions that enable field operational tests on public roads of Remote Automated Driving Systems, and accordingly, the NPA has developed and announced the “Standards for Handling Applications for Permission to Use Roads for Field Operational Tests of Remote Automated Driving Systems on Public Roads” in June 2017.
Under these Standards, field operational tests on public roads of “Remote Automated Driving Systems” may be conducted, with the permission of the NPA for the use of roads. These Standards include, among others:
- the person in charge of testing is required to have a driver’s license. That person may be held responsible in the event of an accident;
- all vehicles are required to be checked for safety at a test course, and to obtain a road-use permit, by having police officers drive in the vehicles and confirm that the vehicles comply with all traffic regulations;
- autonomous vehicle testers are required to inform the local community in advance, and display a message at the front and rear of each vehicle that the vehicle is being tested;
- test permits for autonomous vehicles are valid for up to six months, and tests may only be conducted in areas where there is unbroken wireless access. Tests must also avoid times and places where testing would significantly affect traffic; and
- the autonomous vehicles must be able to be stopped remotely, and have the same level of driving information that a real driver of a vehicle would have.
Summary of the current legal and regulatory landscape
As noted above, there are a number of interacting laws forming a complex legal and regulatory landscape for the operation of autonomous vehicles in Japan.
The following tables provide a summary of the current laws and regulatory landscape for field operational tests and actual operation of automated driving systems on public roads in Japan.
Field Operations and testing
All types of automation
Driver inside the vehicle
No permission required
No driver inside the vehicle (including remote driver)
Permitted by the NPA, if there is a remote driver, and other standards are met.
Level of automation
Level 2 and below
Highly Automated Driving Systems (Level 3 and above, including unmanned)
Driver inside the vehicle
Allowed under the existing laws, and already commercialized
(Traffic-related laws and regulations will need to be revised to allow operation)
No driver (including remote driver)
Roadmaps 2018 specifies the strategies for commercialization of: (i) automated driving systems utilized for private vehicles, (ii) those utilized for business vehicles such as transportations services, and (iii) those utilized for logistics vehicles as an application to the logistics area.
Specifically, the Roadmaps 2018 states that the government will make efforts to realize by 2020:
- commercialization of Level 3 autonomous vehicles that can be automatically operated on expressways,
- Commercialization of Level 2 autonomous vehicles on general roads and
- provision of unmanned autonomous driving transport services (Level 4) in limited areas (e.g., underpopulated areas).
Following 2020, the aims described under Roadmaps 2018 is to realize by 2025:
- commercialization of Fully Automated Driving Systems on expressways;
- popularization of sophisticated driving safety support systems;
- introduction and popularization of automated driving systems in the logistics area, and
- popularization of unmanned autonomous driving transport services (Level 4) for limited areas throughout Japan.
Relevant government agencies have been taking various steps toward the realization of autonomous driving pursuant to Roadmaps 2017, which calls for developing necessary rules, regulations, and policies toward achieving its goals.
By way of example:
- Since August 2017, the Public-Private Council for Automated Driving organized by the Economic Revitalization Bureau of the Cabinet Secretariat
has been holding meetings for managing the progress
of and sharing the results of the public-private partnership-based field operational test projects,
and discussing the necessary institutions for realization of automated driving.
- In parallel, the NPA has also been conducting study and research since August 2017, for step-by-step realization of automated driving based on the direction of the technology development, the aim of which is to consider issues related to the Road Traffic Act.
- Since November 2016, a study group established by the MLIT, the “Study Group on Liability Pertaining to Automated Driving” (the MLIT Study Group), has been considering the liability regime in Japan. In particular, the group has been studying driver and third party liability under the Automobile Liability Security Act, and how such Act should be amended as automated driving increases on Japanese public roads. This is considered further in section D below
- In addition, since October 2016 the Ministry of Economy, Trade and Industry (the METI) has also been conducting a study on the civil liability pertaining to and social receptivity for automated driving. The aim of this study is to consider the gaps between user expectations and technology with respect to automated driving, which party is responsible in the case of an accident, and to examine the social receptivity to automated driving.
- In April 2018, the Comprehensive IT Strategy Office of the Cabinet Secretariat of Japan published the Institutional Development Outline for Automated Driving, which provides a government-wide policy (outline) for the development of institutions for the realization of Highly Automated Driving Systems, including the basic policies with respect to (i) securing safety of autonomous vehicles, (ii) establishing conditions under which autonomous vehicles can operate, (iii) considering appropriate traffic rules,
and (iv) considering civil and criminal liabilities in case of accidents.
Concerned parties from both the public and private sectors are also joining forces to test autonomous vehicles, such as the large-scale highway testing of autonomous vehicles, and the testing of remotely-controllable pilotless vehicles on public roads within designated areas. In the last two years, there have been more than 30 series of field operational tests of automated driving systems conducted on public roads in Japan, led by the Japanese government, local governments, academic institutions, and the private sector.
In the last two years, there have been more than 30 series of field operational tests of automated driving systems conducted on public roads in Japan.
As regards the 1949 Geneva Convention on Road Traffic, under which the actual operation of autonomous vehicles of Level 3 and above on public roads is not allowed, a proposal has been submitted to make an amendment (along similar lines to the 2014 amendment to the 1968 Vienna Convention on Road Traffic) so as to allow autonomous vehicles when automated driving systems can be overridden or switched off by the driver, and work is underway with a view to completing legislation within a few years.2 The expectation is that the relevant domestic legislation (such as amendments to the existing Road Traffic Act) will also be developed in accordance with the amendments to the Geneva Convention. As signatory to the 1949 Geneva Convention, this amendment may potentially impact upon Japanese road laws and regulations relating to autonomous vehicles.
At present, no new laws or regulations for product liability have been made to account for autonomous vehicles in Japan; and product liability relating to a defective vehicle fall under the Product Liability Act of Japan. Below, we describe the main features of that act in the context of autonomous vehicles.
Under the Product Liability Act of Japan (Article 3), a “manufacturer, etc.” of a product is liable for damage arising from the infringement upon another’s life, body, or property which is caused by a “defect” in the delivered product which he/she manufactured, processed, imported, or for which it was represented as their own product.
Nevertheless, Japanese law states that a manufacturer is exempted from the liability under the Product Liability Act if he/she proves:
- The defect in such product could not have been discovered given the state of scientific or technical knowledge at the time when he/she, delivered the product;
- In the case where the product is used as a component or raw material of another product, the defect occurred primarily because of the compliance with the instructions concerning the design given by the manufacturer of such other product, and he/she is not negligent with respect to the occurrence of such defect (Article 4).
The term “defect” as used in the Product Liability Act means a lack of safety that the product ordinarily should provide, taking into account the nature of the product, the ordinarily foreseeable manner of use of the product, the time when the “manufacturer, etc.” delivered the product, and other circumstances concerning the product (Article 2, paragraph 2).
Applying the above laws to autonomous vehicles, should there be a failure of an autonomous vehicle, the car manufacturer which manufactured the autonomous vehicle would likely be primarily liable for the defect in the vehicle.
If the car manufacturer had outsourced the development of the control program in the automated driving system or the supply of necessary data (e.g. high-precision map information), and the program developer or the data supplier had caused the relevant defect, then the car manufacturer may claim damages against the program developer or the data supplier (however, primary liability would remain with the car manufacturer).
Automobile insurance in Japan operates under two different systems and policies: (a) compulsory automobile liability insurance (CALI) and (b) voluntary automobile insurance.
Below, we briefly describe the insurance regime and in addition, consider the regime in the context of autonomous vehicles.
Compulsory automobile liability insurance
Strict liability of an automobile operator under the Automobile Liability Security Act
CALI was established by and operates under the Automobile Liability Security Act of Japan. To ensure financial relief for traffic accident victims, Article 3 of the Automobile Liability Security Act stipulates that any person who operates an automobile for his/her benefit (the “operator”) shall be liable to compensate for death or bodily injury caused to any other person arising from the operation of the automobile, unless he/she is able to prove all three of the following conditions:
- neither he/she nor the driver failed to exercise due care in operating the automobile,
- there was an intention or negligence on the part of the victim or another third party other than the driver, and
- there was no structural defect or functional disorder in the automobile,
where the term “operator” under the Act includes any person who has control over the operation of the automobile and obtains benefit therefrom, regardless of whether he/she owns the automobile or whether he/she is driving or otherwise riding in the automobile at the time of the accident.
This Act transfers the burden of proof of negligence from the victim to the operator of the automobile. As a result, the liability imposed on the operator is very strict, as it is normally not easy to prove all three of the conditions listed above. Article 5 of the Automobile Liability Security Act also obligates every automobile operator to enter into a contract for CALI, in order to secure funds for compensation. As a result, it has become much easier for victims of traffic accidents to be compensated for their losses.
If the traffic accident had been caused by a defect in the automobile, then the automobile operator and its insurer may claim reimbursement from the automobile manufacturer.
Application and issues to consider in the context of autonomous vehicles of Level 4 and above
Whether the operator should be subject to strict liability
The strict liability of an automobile operator under the Automobile Liability Security Act works in the current context of vehicle with a human driver, where statistically, most traffic accidents are caused by driving errors of the driver.
Questions arise in the context of damages arising from an accident of an autonomous vehicle of Level 4 or above. An accident of such a vehicle would likely have been caused by a defect in the vehicle, but the automobile operator and its insurer would effectively be prevented from claiming reimbursement from the automobile manufacturer, since it would be extremely difficult to prove that there had been a defect in the automated driving system. To address this issue, the MLIT Study Group suggested in a report published in March 2018 that new mechanisms to ensure the effectiveness of the operator’s and its insurer’s exercise of their right to obtain reimbursement from the automobile manufacturer, while maintaining the operator’s liability regime;
- to maintain the operator’s liability regime, but introduce mechanisms to ensure the effectiveness of the insurance company’s exercise of its right to obtain reimbursement from the car manufacturer;
- to maintain the operator’s liability regime, but introduce new mechanisms to have a part of the cost shared by the manufacturer in advance, as an insurance premium of the CALI; or
- The suggested new mechanisms include requiring autonomous vehicles to have devices that help to clarify the cause of accidents by recording such information as location, steering and the operational status of autonomous driving systems.
Operator’s duty of care
With regard to the operator’s “due care” (Article 3 of the Automobile Liability Security Act), the operator of an autonomous vehicle at least has the obligation to exercise due care to ensure that the automated driving system does not fail, and to maintain the software and data used in the system (including timely updating of the control system, and not installing programs that are not acknowledged by the manufacturer without its approval).
Whether the operator of autonomous vehicles should have other obligations that are different from the obligations associated with traditional automobiles is a subject of
How to deal with cases of errors in external data, network disruption, and hacking
External data that is used by autonomous vehicles may contain errors. Network communication used by autonomous vehicles may also be disrupted.
Whether traffic accidents by autonomous vehicles due to such errors in external data or network communication disruptions, or due to hacking or other unauthorized influences by a third party into the automated driving systems, constitute a “structural defect or functional disorder in the automobile” (Article 3 of the Automobile Liability Security Act), is a subject of future consideration.
On this issue, the MLIT Study Group’s March 2018 report suggests that an automated driving system that would not operate the vehicle safely in the event of errors in external data, network disruption, or hacking, would possibly constitute a “structural defect or functional disorder in the automobile”
How to deal with own injury caused by a defect in automated driving systems
Under the Automobile Liability Security Act, the CALI of the operator only covers damages caused to “any other person” (Article 3), therefore it does not cover the driver’s or the operator’s own injury.
Whether the operator’s injury arising from accidents is caused by a defect in the automated driving systems should be covered by the CALI is a subject of future consideration.
On this issue, the MLIT Study Group’s March 2018 report suggests that such driver’s or operator’s injury should be covered by the automobile manufacturer’s product liability under the Product Liability Act, the dealer’s tort liability under the Civil Code, and voluntary automobile insurance, for the time being, rather than by revising the current CALI regime.
Voluntary automobile insurance
It is common for insurance companies to offer voluntary automobile insurance products combining various types of coverage, namely, (a) third party liability coverage (i.e., bodily injury liability and property damage liability), (b) self-incurred personal accident coverage, (c) protection against uninsured automobiles, (d) passengers’ personal accident coverage, (e) coverage for damage to the insured’s own vehicle, and (f) bodily injury indemnity coverage (i.e., a wide range of protection against bodily injury suffered by the insured).
The spread of autonomous vehicles is already being anticipated in the world of voluntary automobile insurance in Japan.
At present, voluntary automobile insurance products do not generally cover the accident victim’s damages unless the driver’s fault is established, which is not easy. Determining fault for accidents involving autonomous vehicles can be even more time-consuming and difficult, as automobile manufacturers and technology companies need to be considered. The insurer might not pay a claim if the driver is not established to be at fault, in which case, it would be necessary for accident victims to file damages claims against automobile manufacturers and others, in what is often a protracted process. The possibility of potentially liable parties in hacking cases would make the process even harder, and practically impossible for an individual.
Under such circumstances, in anticipation of the further spread of autonomous vehicles, in April 2017, Tokio Marine & Nichido Fire Insurance Co. added a special condition to all its automobile insurance policies, that stipulates that insurance will be paid out in accidents involving autonomous vehicles
up to Level 3, even if the driver was not at fault.
Since then, other major insurance companies have followed, Sompo Japan Nipponkoa Insurance Inc. in July 2017, and Mitsui Sumitomo Insurance Company, Limited and Aioi Nissay Dowa Insurance Co., Ltd. in January 2018.
Although insurance companies have amended their policies up to Level 3 automation, we anticipate further discussion in the private sector as to how voluntary automobile insurance will be affected by both vehicles up to Level 3 automation, and in the near further, Level 3 automation and above.
Protection of personal information
Data utilization in automated driving systems has been expanding, and it is becoming increasingly important to give due consideration to the protection of personal information and privacy when using data. In particular, the automobile industry has pointed out two major issues in connection with the utilization of a variety of data in automated driving systems:
- securing of consent from individuals to the acquisition of their personal location information, and
- how to handle information concerning surrounding vehicles and pedestrians, which is contained in
The Japanese act relating to the protection of personal information is the Act on the Protection of Personal Information, which was amended and promulgated in September 2015 and fully implemented in May 2017. The act allows private companies to freely use anonymized information (information that has been processed so that no individual will be identifiable from it).
In the context of autonomous vehicles, the examples of methods for processing probe data (information collected by sensors and other equipment, such as location and movement history, that can be accumulated and monitored remotely) are shown in the report published in February 2017 by the Japanese Personal Information Protection Commission, an organization established based on the revised Act on the Protection of Personal Information in January 2016.
In addition, with respect to protection of privacy related to camera images, a guidebook was published in January 2017 (revised in March 2018) jointly by the IoT Acceleration Consortium, the Ministry of Internal Affairs and Communications, and the METI, which describes in detail the matters for business operators to consider when they try to protect citizens and their privacy and communicate with them properly, with a view to promoting utilization of camera images based on their characteristics.
The Japanese government is also considering the risks associated with autonomous vehicles being hijacked by hackers.
Such risks were mentioned in a guidebook on “vehicle information security” primarily aimed at the automotive system industry originally published in August 2013 by the Information-technology Promotion Agency, a government agency of Japan, which was updated in March 2017.
Such risks are also specifically mentioned in the Roadmaps 2018, and are being continually discussed at the MLIT and the study group for automated driving business organized by the MLIT and METI.
At the G7 Transport Minister’s Meeting Declaration (Karuizawa, Nagano, Japan) in September 2016, concerning cybersecurity and data protection, the necessity was recognized for the timely development and regular updating of guidelines and other measures to prevent unauthorized access to vehicles and infrastructure and to protect the privacy of individuals and their personal data. The guidelines on cybersecurity and data protection submitted by Japan and Germany were agreed upon at the meeting of the Subcommittee on Automated Driving of the United Nation’s World Forum for Harmonization of Vehicle Regulations (WP29) held in November 2016, and were subsequently deliberated and adopted by WP29 in March 2017.3
In addition, the Japanese Cabinet Office is planning to create, in FY 2018, its own guidelines on cybersecurity for protecting autonomous vehicles from cyberattacks, to clarify the safety standards.
Since FY 2014, Japan has been promoting public-private partnership-based research and development of automated driving systems, under the Cross-Ministerial Strategic Innovation Promotion Program of the Council for Science, Technology, and Innovation (commonly referred to as the “SIP”), and the SIP has launched industry-government-academia research and development activities on cybersecurity.
In October 2017, the New Energy and Industrial Technology Development Organisation (NEDO), which is a managing entity of “SIP Automated Driving for Universal Services/Field Operational Tests,” selected Deloitte Tohmatsu Risk Services Co., Ltd., Nihon Synopsys G.K., and PwC Consulting LLC to conduct the tests in which a vehicle’s resistance to mock cyberattacks will be tested. This testing will be part of a series of field operational tests of autonomous vehicles that the Cabinet Office is administering on public roads, involving 21 automakers and other groups, to be conducted through March 2019. First, a method of appraising autonomous vehicles’ resistance to hackers will be established.
From summer 2018, the autonomous vehicles will be put under mock cyberattacks in research facilities, and tested using the established appraisal method. It is contemplated that the results from the test will be incorporated when the Cabinet Office compiles the guidelines on cybersecurity in FY 2018. The Roadmaps 2018 also states that, in order to strengthen security measures, it is important to reinforce the a system operated by the Japan Automobile Manufacturers Association since April 2017 for sharing information on responses to incidents among companies (a Japanese equivalent of the
- The term “operational design domain (ODD)” means the specific conditions under which the driving automation system is designed to function, including, but not limited to, driving modes.
- Since the drafting of this paper, two proposals to amend the convention did not come to have effect.
- Since the drafting of this article, we understand that there have been further adopted proposals to WP29. Updates will be provided shortly by separate news alert.