DSGTas - Department of State Growth - Tasmania23 Mar 2016
Thank you for giving the Tasmanian Department of State Growth the opportunity to comment on the NTC issues paper - Regulatory barriers to more automated road and rail vehicles. These technologies have significant potential to improve both road safety and traffic flows to reduce congestion.
The issues paper provides a comprehensive overview of the areas that policy makers should focus on when considering a coordinated national approach to the development of automated vehicle technology. Below are some further issues from a Tasmanian perspective, in response to your specific consultation questions.
What are automated vehicles?
Question 1 – Do you support the use of the Society of Automotive Engineers (SAE) International Standard to classify automated road vehicle functions? Do you have any issues with using the SAE International Standard?
In principle – yes. With technologies progressing as swiftly as they are, it’s difficult at present to know whether it’s an appropriate classification schedule. It seems to encompass all technological possibilities but will it tie regulatory authorities into certain legal and technical definitions in the future that may not be appropriate? It seems suitable for now, but we suggest this will need regular monitoring and revision as the technology and regulatory environment matures.
Role of government
Question 2 – What do you think the regulatory role of governments should be to support the introduction of automated vehicles in Australia?
The regulatory role of governments is in reviewing existing regulation, identifying areas that are inapplicable to automated vehicles or that serve as unnecessary regulatory barriers to their implementation. Regulators will have to work extensively and with agility - with each other, and with industry - and keep up to date and continuously fine-tune our regulatory response as these technologies evolve.
Issues with regulating the driver
Question 3 – Have we identified the key issues relating to the Australian Road Rules and state and territory road safety and traffic laws? Are there other issues that should be assessed as part of the NTC review?
Currently in the Tasmanian Road Rules, Rule 16 defines the driver as a “person”, so they are explicitly human. Redefining “driver” may get around most existing legislative issues, allowing the driver to be human or system, while specifying where a provision applies to a human but not to a system (e.g. drink driving), and vice versa. In reality most of our existing laws will need to apply as it will be long-term before there are no “human” drivers.
Discretionary rules such as “when safe to do so”, “when practicable” and “reasonable” (e.g. reversing for a reasonable distance) will undoubtedly cause a problem for automation technology, and how developers choose to overcome these issues may shape how our road law looks in the future. In the same vein, there are also many “unspoken” rules that form part of our behaviour on the roads – letting someone in when merging or changing lanes, giving way to a pedestrian when not required to by law, letting someone else go even if it’s your turn just to help traffic flow. While these unspoken rules may not form part of the regulatory response, they may need to be factored in (to some extent) by system designers, to maximise safety and public acceptance.
The introduction of fully automated vehicles raises the questions of whether a person would still need to be physically/mentally fit in the event that something did go wrong and they had to take control. This may have an effect on the national standards for medical fitness.
There will need to be some link between existing licence holders and their ability to monitor/control an automated or semi-automated vehicle. If the technology goes down the path of a “driver” needing to monitor the task and intervene if needed, how will we know that person is capable of doing so? The competency of someone learning to drive in an automatic vehicle may be very different to someone whose driving career has been in traditional vehicles. Would we require re-licensing or re-testing for the approved use of some vehicles, or trust that drivers will adequately inform themselves of their vehicle’s capabilities prior to driving as we do now?
Intervention in a car (human-to-system and system-to-human) is also likely to be needed within a fraction of a second to avoid an incident, depending on the circumstances. Will the warnings and changeover abilities of the automation system be adequate for the realities of swapping between automated and manual driving? As well as being a technical issue for system designers, this is also a training and awareness issue for regulators in their role of ensuring the competency of drivers.
There is also the issue of mixed technology (manual override capability and requiring the driver to take control in some instances) and how we ensure regulatory compliance and prove a case when we may not know for certain who is in control of the vehicle. For example, in those vehicles where manual control is still an option, speed laws should still apply. There may also be instances where a fully automated vehicle is programmed to exceed the speed or break a road rule in order to avoid a collision – if this information was recorded in a telematics device, we would need mechanisms to enable enforcement bodies to obtain and take into account evidence of extenuating circumstances.
Issues with regulating the road vehicle
Question 4 – Have we identified the key issues relating to the Australian Design Rules and other vehicle standards? Are there other issues that should be assessed as part of the NTC review?
Whilst we might have some specific local design requirements, it’s likely that international technological advancements and vehicle requirements and safety standards will be similar to Australia’s.
At the moment there is a fairly clear delineation of responsibilities between ADRs (Commonwealth) and Road Rules and Vehicle Standards (states and territories - although using nationally agreed model law). Whilst the states have representative involvement in the ADR development and approval process, if the role of ADRs and thus the Commonwealth is to increase, there may need to be new, more suitable governance structures in place, i.e. reconsidering the role of the states. We already have dual vehicle standards for heavy vehicles and light vehicles, an added layer of automated vehicles may see the need for a rethink of the vehicle standards regulatory process altogether.
With a much bigger reliance on the vehicle and in-vehicle systems, should we legislate, or provide some other mechanism, to ensure that vehicles are kept in good working order (e.g. regularly serviced and operating systems are regularly updated and checked)? A closely related issue is ensuring local technical expertise is available for monitoring and servicing of more automated vehicles to the standards needed, including considering the levels of accessibility to these services for residents in more rural or remote locations.
Regulation and considerations of a move to a risk based approach (e.g. rail safety) raises issues around enforcement. The volume of driver-related infringements may reduce, however the complexity of remaining infringements may increase. Also, with traditional road/traffic regulation a person can be issued an infringement and will be deemed liable if they take no further action. Should we move in future to a positive obligations model, we would see a much greater need for prosecutions staff and a greater impost on courts.
It also raises questions about who will have a case mounted against them in different circumstances – e.g. in the case of full automation would it be vehicle manufacturers or those responsible for administering the IT systems used by vehicles and infrastructure? What evidence will we use then? Would the vehicle occupant (“driver”) be a witness for or against the prosecution? Those bodies will also likely be larger corporations with resources to mount a lengthy defence.
As the issues paper notes, more automated vehicles will need to be regulated alongside our existing vehicle fleet long term. With vehicle standards it may be relatively easy to operate an exemption process concurrently with our existing processes, but this will become more difficult when we explore the interplay of driver, vehicle, road/traffic rules and enforcement officers.
Issues with regulating heavy vehicles
Question 5 – Have we identified the key issues relating to heavy vehicles? Are there other issues that should be assessed as part of the NTC review?
Having two different regulatory frameworks for light vehicle standards and heavy vehicle standards will add another layer of complexity to any plans to regulate automatic vehicle use in Australia. There is also an additional regulatory body to consider in the National Heavy Vehicle Regulator (NHVR).
Question 6 – Have we identified the key issues relating to the liability of drivers, manufacturers, service providers and road managers? Are there other issues that should be assessed as part of the NTC review?
As the technology changes and focus progressively shifts from the driver doing the right thing to the vehicle and road infrastructure doing the right thing, there will be a commensurate shift in liability.
There is a risk that road authorities could be held liable (as a party) in the event of a crash in an automated vehicle if the infrastructure does not meet the relevant standards. This could cause increased costs to the State to: a) upkeep and improve the infrastructure; and b) damages awarded against them. If such a liability risk exists, road authorities may be forced to take a very conservative approach to the types of infrastructure and environments that they will permit more automated vehicles to travel on.
Under a risk based approach and positive duties, the onus would likely be on the vehicle manufacturer to ensure that the person knew how to adequately control the vehicle. Thus, if the person caused a crash or offence by failing to intervene or take control when prompted, would the vehicle manufacturer be held accountable, or the driver?
Privacy and access to data
Question 7 – Have we identified the key issues relating to privacy and access to data by government agencies? Are there other issues that should be assessed as part of the NTC review?
The commonwealth and states currently have a solid framework around privacy and access to data. This may need some adjusting in line with technological developments, but the collection and use of in-vehicle data is already regulated and it would not likely present any insurmountable barriers to vehicle automation.
Supporting on-road trials
Question 8 – Have we identified the key issues relating to on-road trials of automated road vehicles? Are there other issues that should be assessed as part of the NTC review?
As we have seen with South Australia and Western Australia, as well as internationally, on-road trials will be the first step towards fully implementing automated vehicles. So far, jurisdictions seem to have been able to adjust their legislation or introduce exemptions to allow for trials to occur.
While there hasn’t been much coordination amongst jurisdictions at this stage, the advantage of allowing differing approaches early on is that there will be “testing grounds” in different jurisdictions that are hopefully well suited to local conditions and allow for comparisons and evaluations of the different trials. A practical issue may be the lack of infrastructure in some smaller jurisdictions to separate automated and conventional vehicles in the earlier trial stages. If so, it may not be possible (or necessary) to recreate all trial phases in each state or territory.
More automated rail
Question 9 – Have we identified the key issues relating to more automated rail operations? Are there other issues that should be assessed as part of the NTC review?
As automated and semi-automated rail is already (or soon to be) a reality in most jurisdictions, consideration of the issues around automated rail vehicles are likely to be different to road vehicles, as there is more limited infrastructure and little opportunity to mix with other users such as road vehicles and pedestrians. Rail safety is also well designed for future developments in that its focus is on risk mitigation and positive safety obligations that are generally flexible and adaptable.
The laws and infrastructure developments around rail crossings and train stations will need to be a major consideration, as these would be the likely points of conflict with people and other vehicles.
There are probably lessons to be learned from both the rail and aviation industry when looking at road vehicle automation, as they both have well-developed legislative and procedural systems in place to manage automation issues.
Question 10 – Are there additional issues or risks that should be considered in the NTC’s assessment of regulatory barriers to more automated vehicles?
As a smaller jurisdiction, Tasmania has some unique issues to consider when looking at the introduction of automated vehicles. We have traditionally had an older vehicle fleet than most jurisdictions, so it is likely that this would continue and we would have less automated vehicles and would foreseeably maintain a mix of automated and traditional vehicles for longer than most jurisdictions.
Smaller jurisdictions obviously face ‘critical mass’ issues in developing an appropriate regulatory framework and providing (as road authorities) the necessary supporting infrastructure. Any national programs for implementation will need to pay serious heed to these resourcing constraints and maximise opportunities for national and multi-state collaboration.