Conduct of the inquiry
The Industry, Innovation, Science and Resources Standing Committee launched its inquiry into the social issues relating to land-based driverless vehicles on 1 December 2016. The Terms of Reference are available on page vii of this report.
The Committee’s focus throughout the inquiry was on the social aspects of driverless vehicles in Australia. While technological developments and questions and regulatory approaches were considered within that context, the Committee’s priority was to establish whether the social changes likely to be brought about by the introduction of highly automated vehicles in Australia are being adequately considered.
The Committee received 47 written submissions and held ten public hearings and four inspections between March and June 2017, in Canberra, Perth, Melbourne, Brisbane and Sydney. In total, the Committee heard from more than 30 witnesses. A list of submissions received is at Appendix A. The names of witnesses and a list of hearings are at Appendix B.
This report, transcripts of public hearings and submissions are available on the Committee’s website: www.aph.gov.au/iisr.
Definitions and key terms
The Committee heard that there is considerable debate regarding the preferred terminology for driverless cars. While ‘driverless vehicles’ is the phrase most-readily understood and recognised, several witnesses and submitters argued that it is misleading and potentially off-putting to members of the public.
The National Transport Commission explained the rationale for using the term ‘automated’:
We have used the term 'automated' rather than 'driverless' or 'autonomous' to recognise that it is a spectrum of automation and that there are different policy issues as we move along that spectrum. We are already on that journey today with today's cars.
The Department of Infrastructure and Regional Development (DIRD) uses the term ‘automated vehicle’, meaning ‘a vehicle that does not require a human driver for at least part of the driving task’ and notes that this term captures a broader range of vehicles than the more specific descriptor ‘driverless’.
In contrast, the Australian Driverless Vehicle Initiative (ADVI), a peak advisory body comprised of a range of organisations across government, academia and industry, noted that:
We use the term 'driverless' because we had a focus group and a number of discussions with our key people about what language the community would understand and relate to and, whilst our technological experts like 'automated', 'driverless' is the language that we use. We refer to everything as 'driverless' across the spectrum.
This report uses the terms ‘driverless’, ‘automated’’ and ‘autonomous’ interchangeably, generally seeking to maintain consistency with the evidence being referred to.
Levels of automation
The standard definition for the automation level of vehicles is the one used in the Society of Automotive Engineers’ (SAE) International Standard J3016. This definition has been adopted by authorities in Australia, Europe and the United States of America and was consistently used by submitters and witnesses to this inquiry.
Under the SAE standard, there are five levels of automation. All vehicles will fall into one of these categories:
Table 1.1: SAE Standard Levels of Automation
Human driver responsible for all aspects of the driving task.
In some circumstances the system is capable of either steering or acceleration/deceleration (including braking), with the expectation that the human driver performs all remaining aspects of the driving task.
In some circumstances the system is capable of both steering and acceleration/deceleration. The human driver must monitor the driving environment and respond as needed.
Level 2, but when the system is operating in automated mode the human driver is not required to monitor the driving environment. The human driver must respond to requests from the driving system to intervene.
Level 3, but no human monitoring or intervention is required when the system is operating in automated mode.
Automated system in control all of the time, and in all road environments.
Source: Department of Infrastructure and Regional Development, Submission 26, p. 7.
As discussed above, the term ‘driverless vehicles’ therefore only strictly applies to vehicles meeting the characteristics of Levels 4 and 5 of the SAE Standard. Vehicles meeting the characteristics of Levels 1 and 2 are already commercially available in Australia and used on public roads; vehicles at higher levels are used only in controlled environments, including mining sites or specific trials including the RAC Intellibus in Perth, which the Committee inspected, and the Darwin Waterfront bus.
Submitters and witnesses to this inquiry expressed a range of views regarding when vehicles with high levels of automation can be expected to be commercially available for use on public roads.
Manufacturer Volvo, for instance, noted that public perceptions of driverless vehicles are probably based on inaccurate understandings of the current state of technology:
Public confusion is exacerbated by regular media reporting which describes a utopian world in which drivers are transported from home to office in fully driverless, ‘handsfree’ vehicles. Based on future technology advances this scenario is certainly achievable in Australia, but realistically it is probably some decades away.
Vehicle technology supplier Robert Bosch (Australia) provided the below image to illustrate the likely timeframe of technological development, showing increasing levels of assisted driving over the next few years and highly automated driving after 2020:
Figure 1.1: Robert Bosch Technology Roadmap
The NTC noted its timeframe for regulatory reform – as endorsed by the Transport and Infrastructure Council in November 2016 – is based on the following assumptions developed after consultation with industry:
demand to trial different levels of driving automation on public roads is already occurring and is expected to increase significantly in the next two to three years.
large-scale commercial deployment of increasingly automated vehicles that still require a human driver is expected by 2020.
large-scale commercial deployment of automated vehicles that do not require a human driver (for some, or the entire journey) is expected after 2020.
Models of car ownership
A point made by many of the witnesses and submitters to the Inquiry was that the introduction of driverless passenger vehicles can, and likely will, lead to a shift in the model of car ownership for most Australians. Opinions varied amongst the Committee’s witnesses as to the extent of the likely change – ranging from those who believe there will be only minimal change through to those who argued that private car ownership will become a thing of the past. However, most witnesses agreed that there will be at least some impact on car ownership as driverless passenger vehicles become available.
Furthermore, most witnesses argued that many of the anticipated benefits of driverless vehicles will become most apparent if there is also a substantive change in the way that Australians own and use cars. DIRD’s submission, for instance, noted that a model based on less ownership and more ride-sharing would have positive effects including ‘improving mobility in Australia’s densely populated urban areas and providing an incentive for travellers to move away from private car ownership and make better use of public transport’.
Other witnesses noted that the early signs of this shift away from private ownership are already appearing: young Australians are choosing to get their licence later than previously, and car ownership, particularly in inner-urban areas, has fallen.
Professor Bradlow of the Australian Academy of Technology and Engineering, for instance, argued for a strongly interventionist approach to ensure that the anticipated benefits of autonomous vehicles transpire, suggesting that governments should choose ‘to make individual ownership prohibitively expensive, because it is no longer necessary at that point’.
DIRD’s submission also pointed to a model wherein autonomous vehicles remain privately owned at about the same rate as vehicles currently are, wherein ‘a number of perverse outcomes might occur’. These could include increased congestion (as vehicles drop passengers off and return to their residence, as well as by making longer commutes and reducing the attractiveness of public transport options) and possible negative environmental impacts.
Structure of the report
Each chapter of this report addresses one of the Inquiry’s terms of reference, with the final chapter focusing on the role of government in preparing for driverless vehicles.
Chapter 2 provides an overview of current public attitudes towards driverless vehicles, drawing on existing surveys and studies. The chapter discusses the major identified benefits of and concerns about driverless vehicles, and introduces the main theme of this report: that public engagement is of paramount importance. The concerns of the public will need to be addressed if the benefits of autonomous vehicles are to be realised in Australia. Similarly, stakeholders – including governments and manufacturers – will need to educate and emphasise the benefits which these vehicles can bring.
Chapter 3 focuses on the first of those benefits, and the one most anticipated by submitters and witnesses to this inquiry: the improved safety outcomes which driverless vehicles are likely to bring. Noting that human error is responsible for the overwhelming majority of road trauma, the capacity of driverless vehicles to significantly reduce the likelihood of accidents is a substantial benefit. The chapter considers safety for both passengers and non-passengers, as well as identifying some of the unresolved ethical concerns which driverless vehicles will bring.
Following on from safety, chapter 4 discusses one of the key concerns identified in the research to date: the lack of clarity around questions of legal liability and responsibility if an autonomous vehicle is involved in an accident. Recognising that a fully autonomous road system will see a significant change in terms of road safety, the chapter also addresses what changes can be expected in the insurance industry as a result.
Another key concern identified is the shift in employment that autonomous vehicles will bring. Chapter 5 discusses the main issues in relation to this question, including noting that the impact will be wider than the immediately affected sectors such as professional drivers. Chapter 5 concludes by noting some of the opportunities for Australia in new employment fields brought about by the introduction of autonomous vehicles, and the skills required to take those opportunities.
Chapter 6 is based on another one of the most important benefits expected of driverless vehicles: the increased access and mobility they will offer, particularly to people with disability or older Australians no longer able to drive. By opening up transport possibilities, autonomous vehicles may serve to significantly improve the lives of many Australians. As discussed in the chapter, however, it is important that this new technology does not serve to widen the divide between those who have the means to access driverless vehicles and those who do not.
Another aspect of travel likely to change substantially with the availability of autonomous vehicles is public transport systems. Chapter 7 addresses this topic, including the prospect that driverless technology will greatly expand access to public transport by leading to entirely new public transport applications. Concerns and possible issues with driverless public transport options are also discussed.
The final chapter outlines some of the roles governments will need to fill to enable the introduction of land-based driverless vehicles in Australia. These include ensuring consistency in regulations and standards across the country and working towards consistent and adequate infrastructure. It also includes leading public dialogue on the many social impacts Australia is likely to see as driverless vehicles become available. Public acceptance and engagement will be necessary to ensure that the anticipated benefits of driverless vehicles will be seen in Australia, and governments have a central role to play in ensuring that this engagement occurs.
The Committee recommends that the Commonwealth adopt as standard terminology the use of ‘automated vehicles’ and formally accept that the standard definition for the automation level of vehicles is that used by the Society of Automotive Engineers’ (SAE) International Standard J3016. The Committee recommends that the use of ‘driverless car/vehicle’ and ‘autonomous vehicle’ be discontinued.