Florida puts autonomous vehicles in the fast lane

Florida continues to lead the way in state level regulation of autonomous vehicles; recently passing legislation that permits autonomous vehicles and on-demand autonomous vehicle networks on public roads.

HB 311: Autonomous Vehicles, sponsored by Rep. Jason Fischer, and SB 932 Autonomous Vehicles, sponsored by Sen. Jeff Brandes, have sailed through Florida’s legislature and only require Governor Ron DeSantis’ signature to officially pass into law. The House-Senate bills seek to provide uniformity of laws governing autonomous vehicles, thereby clearing the regulatory path for autonomous vehicle development and increasing investment in the Sunshine State.

The legislation expressly permits fully autonomous vehicles to operate in Florida regardless of whether a human operator is physically present in the vehicle. “Automated driving systems” and “fully autonomous vehicles” are amongst the key terms defined within the legislation that exempts autonomous vehicles and operators from certain prohibitions (including on the active display of television or video and the use of wireless communications devices systems while driving).

SB 932: Autonomous Vehicles authorises the Florida Department of Transportation, in consultation with the Department of Highway Safety and Motor Vehicles, to conduct pilot or demonstration programs to explore the efficient implementation of innovative transportation technologies.  It also authorises the Florida Turnpike Enterprise, a business unit of the Florida Department of Transport, to enter into one or more agreements to fund, construct, and operate facilities for the advancement of autonomous and connected innovative transportation technologies for certain purposes.

The legislation provides requirements for insurance and operation of on-demand autonomous vehicle networks, such as Uber or Lyft, and revises the registration requirements for autonomous vehicles. SB 932, legalises “on-demand autonomous vehicle networks,” defined as a service that uses “a software application or other digital means to connect passengers to fully autonomous vehicles, exclusively or in addition to other vehicles, for transportation, including for-hire transportation and transportation for compensation.”

Florida is widely regarded as one of the leading states in the development of self-driving vehicle policy. Fischer, the architect of the House Bill, stressed the importance of this legislation in maintaining Florida’s position: “To maintain this position and encourage companies to test and deploy in our state, we must address our existing laws governing motor vehicle operation that did not contemplate a driverless future when they were written.”

Perhaps unsurprisingly, these developments have been positively received from the likes of Uber. Senior policy manager, Stephanie Smith, commented that: “This measure provides direction on the roles of state and local government and authorization for the deployment of automated vehicles on a ride-sharing network. These provisions establish a clear pathway to bring the benefits of automation to our state.”

Regulatory advancements of this kind are a necessary precursor to the deployment and commercialisation of autonomous vehicles.  While this is a welcome development for Florida, the barriers to entry and upscaling created by this ‘state-centric’ approach to regulation continue to threaten the United States’ global leadership position in this sector.

Joseph Falcone
Joseph Falcone
Partner, New York
+1 917 542 7805
James Allsop
James Allsop
Senior Associate, Tokyo
+81 3 5412 5409
Peter Keeves
Peter Keeves
Graduate Solicitor, Australia
+81 3 5412 5427

 

Who is liable when a connected and autonomous vehicle crashes?

In March 2018, a pedestrian in Arizona was struck by a self-driving Uber vehicle during a vehicle test. An Arizona prosecutor recently found Uber not criminally liable for the incident, though little detail was given for the decision (see our blog post, here).

Once CAVs are on the roads and self-driving, liability is expected to shift away from the driver towards the CAV itself. Liability for accidents should, therefore, lie with manufacturers. This article explores such manufacturers’ liability for on-road CAV accidents under the current Australian regime .

CAVs and the product liability regime

CAV-related liability will arise against a potentially complex factual background involving a multi-part supply chain, the highly technical interplay of parts and processes, varying states of control between driver and vehicle and the likelihood of continuous software updates to CAVs post-purchase. However, Australia’s existing product liability regime has proven to be adaptable across many industries (including medical devices, car parts, consumer goods and pharmaceuticals).

A potential product liability claim may have several bases under current law. The cause of action pursued will be influenced by the injured party’s identity –  whether a driver / passenger or third party. A party injured in a CAV crash may argue:

  • negligence: that a defect in or failure of the CAV breached a duty of care owed to the claimant by the manufacturer and / or supplier;
  • breach of contract: that a CAV failed to operate in accordance with the terms of a contract between a consumer and manufacturer;
  • safety defect (Australian Consumer Law (ACL)): that injuries were suffered due to the CAV having a ‘safety defect’. A defective goods action does not require a contractual relationship between the injured party and the manufacturer;
  • breach of statutory guarantees (ACL): that a CAV was defective and therefore its supply breached statutory guarantees (eg vehicles are of ‘acceptable quality’ and are ‘reasonably fit’ for their disclosed purpose); and / or
  • misleading or deceptive conduct (ACL): that a manufacturer and / or supplier’s representations to consumers (eg the CAV was safe for use) were misleading or deceptive.

The ACL-based claims demonstrate both the regime’s flexibility and certain challenges in bringing a CAV-related product liability claim.

(a) The ‘manufacturer’ and the ‘good’ 

‘Manufacturer’ has a broad meaning under the ACL, capturing the actual manufacturer, someone who holds themselves out as the manufacturer and the good’s importer. The term ‘good’ is similarly broad, and includes ‘vehicles’, ‘computer software’ and ‘any component part of, or accessory to, goods’. This arguably covers all aspects of the CAV, such as sensors, cameras, navigation software and data transmission systems.

Future car ownership and use norms will also play a role. In the increasingly popular pay-per-use model, a consumer may not be acquiring a good but instead receiving a service. The ACL captures service provision, with mirror statutory guarantees and misleading or deceptive conduct provisions.

(b) Who should have been driving?

Different levels of automation will require different levels of driver input. Establishing who should have been driving when a crash occurred will have significant implications for legal causation.

While the current legal regime could likely address such questions in a piecemeal way, through existing obligations regarding instructions for use and the usual principles of evidence, future regulation on fundamental issues such as ‘control’ will likely assist in consistent and predictable legal outcomes.

For example, future regulations may require that drivers be in control in high pedestrian areas, notwithstanding that a CAV is designed to function in those environments. The National Transport Commission (NTC) is considering this (see the NTC’s relevant policy paper, here). The question of control would fall away for ‘fully autonomous’ CAVs – although we query if a manufacturer would ever make such an unqualified statement.

(c) What is the defect?

Numerous factors specific to CAVs could conceivably cause accidents, including network outages, computer viruses, unpredictable behaviour by other road users and road maintenance conditions.

An accident may occur with no clear ‘defect’ and the CAV functioning exactly as intended. The Uber crash in Arizona appears to have been caused by the purposeful disabling of the ‘emergency braking manoeuvre’ system, disabled to reduce the potential for erratic vehicle behaviour during computer-controlled driving.

Separately, the ACL provides a defence to manufacturers where there was no defect at the time of supply. Consider the application of such a defence where, for example:

  • a crash is caused by a temporary network failure; or
  • a defect arises post-supply which can be rectified with a software or hardware upgrade.

A plaintiff may need to commence proceedings against a broad range of defendants to ensure recovery can be made against the defendant actually responsible. Large-scale changes to the legal liability and insurance regimes for CAVs may be necessary to make such proceedings more cost-effective.

(d) What instructions were given?

Instructions given to consumers by CAV manufacturers will be of paramount importance – especially those relating to maintenance, driver / CAV control sharing and driver behaviour during CAV-controlled driving. Consider, for example, a scenario where the manufacturer’s instructions require consumers to regularly install software updates to the CAV. If a crash was caused by a glitch which the software update would have corrected, questions arise around compliance with instructions and the steps taken by the manufacturer to ensure updates are made (eg should a manufacturer disable the CAV until updates have been installed?)

Conclusion

The Australian product liability regime seems largely able to handle the novel manufacturer liability claims which may arise from the widespread introduction of CAVs. Fundamental issues impacting liability, such as ‘control’, will no doubt be addressed as part of the safety risks associated with CAVs (see the NTC’s relevant discussion paper, here). As Courts identify issues over time, further legislative reform around the liability and insurance regimes will likely be desirable, taking into account any industry guidance and best practice developed.

For now, we continue to monitor, with great interest, how technological, regulatory and commercial progress will shape consumer consumption of the CAV.

Guy Narburgh
Guy Narburgh
Special Counsel, Australia
+61 2 9322 4473
Camilla Pondel
Camilla Pondel
Solicitor, Australia
+61 2 9225 5835

Inquiry into Automated Mass Transit

Following a five month inquiry into automated land-based mass transit, House of Representatives Standing Committee on Infrastructure, Transport and Cities released its Innovating Transport Across Australia Report in March that in essence recommends the Government needs to lead the transport revolution. The purpose of the inquiry was to report on current and future developments in the use of automation in land based mass transit, including the role and responsibility of the Australian Government in developing suitable technology.

Major players in the connected and autonomous vehicle industry made submissions to the inquiry and the key themes raised in the submissions included:

  • the importance of a nationally consistent legal and regulatory framework;
  • ensuring the laws and regulatory framework allows for the safe deployment of technologies that are required for the automated land-based mass transit; and
  • equipping Australian infrastructure to accommodate new automated and electric vehicle technologies.

The Committee sees the automation and electrifications of mass transit as having the potential to make Australian cities and regions cleaner, greener, more accessible and more liveable. Committee Chair John Alexander says achieving this outcome will require Australian Government’s leadership and vision.

5 key recommendations from the report:

  • Support the development of a new automated transport ecosystem

One of the biggest challenges for automated mass transit is the problem of the first and last mile – the gap between the mass transit services and the passengers’ home or destination. An automated transport ecosystem is envisioned to involve autonomous rail, light rail and buses providing for the high volume routes, which are then connected with smaller vehicles like shuttle buses or driverless cars that can provide connectivity to the passengers’ first and last mile. The development toward a new automated transport ecosystem requires a collaborative effort from the Australian, State and Territory Governments to introduce adequate policies and regulations that support such an ecosystem.

  • Developing a national hydrogen strategy

The Committee supports the development of a national hydrogen strategy. The Australian Government Chief Scientist proposed a national hydrogen strategy in December last year to capitalise the emerging economic prospects of hydrogen as an energy source. The Committee sees hydrogen based technologies playing a significant role in the future of Australian land transport however, hydrogen power brings its own infrastructure demands and the challenge is to identify the optimum pathway. The Committee recommends that the Australian Government work collaboratively with the State and Territory Governments to develop a national hydrogen strategy that provides for the manufacture and transport of hydrogen in a safe, cost effective and energy efficient way.

  • Facilitate the uptake of electric vehicles

The Committee recommends the Australian Government to increase its uptake of electric vehicles, both Battery Electric Vehicles and Fuel Cell Electric Vehicles. Electrification of transport has real potential to lower costs, reduce the environmental impact of land transport and enhance national fuel security. To ensure the electric vehicles are successfully introduced to Australia, the Government is encouraged to develop appropriate infrastructure required for the deployment of electric vehicles, ensure refueling and recharging technology are compatible with defined standards and promote greater coordination between the transport and energy sectors.

  • Undertake research to estimate the national requirement for electricity generation under an electric and automated transport future

The Committee recommends the Australian Government to estimate the national requirement for electricity generation if Australia moves to an electric and automated transport future. This ought to be done with a view to ensure that the electricity generation will meet the anticipated demand, whilst also being environmentally friendly. Namely, the Committee recommends that the Government adheres to the national greenhouse gas abatement targets, when meeting the national electricity needs.

  • Establishing statutory offices and expanding the role of current statutory bodies to facilitate the automated land-based mass transit future.

The report highlights the need to rethink the current operation of statutory bodies to enable them to play a more productive and significant role in enabling automated land-based mass transit.

The Committee specifically recommends the establishment of the statutory Office of a National Chief Engineer to provide independent expert advice on the planning and development of Australia’s infrastructure.  The Department of Infrastructure, Regional Development and Cities is recommended to conduct an audit of Australia’s existing transport communications infrastructure and requirements for automation at various stages. The audit is to help develop a national strategy for transport communication infrastructure for full automation of land transport. The Committee is also of the view that given the nexus between automation and electrification of transport, the Office of Future Transport Technology, which sits within the Department of Infrastructure, Regional Development and Cities, to expand to cover alternative energy sources such as battery electric power and hydrogen fuel cell power.

Susannah Wilkinson
Susannah Wilkinson
Senior Associate, Brisbane
+61 7 3258 6786
Stebin Sam
Stebin Sam
Graduate, Brisbane
+61 7 3258 6315

TOYOTA’S GUARDIAN SYSTEM: A MIDDLE ROAD BETWEEN MAN AND MACHINE?

At the 2019 Consumer Electronics Show in Las Vegas, Toyota made two revelations.  First, it revealed “Guardian”, an automated driving safety system that it has developed in parallel to its autonomous driving system, “Chauffeur”.  Second, it revealed a desire to share this technology with other auto manufacturers.

It is not every day that a manufacturer announces a desire to share its technology with its competitors.  Why does Toyota want to do this?  Its answer was clear: to try and save as many lives as possible.  How does Toyota intend to do this?  That seemingly remains to be decided.  In making that decision, there are a number of potential legal issues that may be relevant.  In this bulletin, we consider some of these issues as well as the question of whether this technology may present an alternative to fully autonomous vehicles.

What is the Guardian system?

Guardian is a passive safety system that is designed to assist in controlling vehicles which are otherwise controlled by a human driver at all times.  It is not a fully autonomous system that seeks to replace the human driver.  Rather, it is designed to augment human control of a vehicle, including by intervening in dangerous scenarios.

How the Guardian system works

Toyota has billed Guardian as the application of its studies of ‘fly-by-wire’ control systems in fighter jets.  But what does this mean?

Traditionally, fly-by-wire meant that the signals created by a pilot’s interaction with the flight controls were converted to electronic signals which passed to the aircraft’s flight control computers, thereby replacing manual flight controls.  However, in advanced systems this has come to mean that, rather than responding directly to the pilot’s inputs, the flight control computer seeks to identify the pilot’s intended objective and then triggers outputs aimed to achieve that (subject to remaining within a set safety performance envelope).

With Guardian, Toyota is trying to apply this system to cars.  However, unlike in a fighter jet, the safety envelope is defined not only by the parameters and performance of the vehicle, but also by the vehicle’s perception and prediction of everything in its immediate environment, including pedestrians and other vehicles.

An alternative to fully autonomous vehicles?

Toyota does not see Guardian as merely a stepping stone technology while fully autonomous driving systems are developed.  Rather, it regards it as an important alternative; a system that allows people to continue to drive but with the benefit of as-needed technological assistance.

This raises an interesting question regarding the potential for long term mixed fleets on public roads.  A number of commentators have suggested that the proliferation of autonomous vehicles will sound the death knell for conventional vehicles; conventional vehicles will either (a) become prohibitively expensive to insure, or (b) be banned on safety grounds.  As a result, while there will likely be a period of mixed fleets on public roads, conventional vehicles will eventually die out and autonomous vehicles will become ubiquitous.

By introducing an automated safety system that augments the human driver, could Guardian provide the best of both worlds: allowing humans to continue to drive with a technological safety net to prevent accidents?  Time will tell.  However, if the Guardian system is able to demonstrate safety benefits comparable to that expected of fully autonomous vehicles, it may be possible to imagine a scenario where there is a long term mixed fleet of fully autonomous vehicles (for those that do not wish to drive) and “augmented” vehicles (for those that do).

Potential anti-trust issues

Despite the potential benefits for drivers, Guardian is not being developed with only drivers in mind.  Toyota’s aim is that Guardian will be capable of operating not only with a human driver, but also with any autonomous driving system, irrespective of the manufacturer.

To realize this aim, it will be crucial for Toyota to ensure that its system is interoperable with other autonomous vehicle technologies.  In order to achieve this, a level of standardization may be required, which can potentially raise questions of competition law.

Generally, standardization is considered to have positive economic effects and governments have been reluctant to restrict it.  However, where standardization risks creating market power (effectively the ability to profitably maintain prices above or output below competitive levels) this can be more problematic.  In recognition of this, the European Commission specified criteria which, if met, allow a standardization agreement to benefit from safe-harbor against being deemed anti-competitive under Article 101 of the Treaty for the Functioning of the European Union.  These are as follows:

  • participation in the standard-setting is unrestricted;
  • the procedure for adopting the standard is transparent;
  • there is no obligation to comply with the standard; and
  • companies have effective access to the standard on fair, reasonable and non-discriminatory (“FRAND“) terms.

While failure to meet these criteria will not automatically render a standardization agreement anti-competitive, care will need to be taken by manufacturers to ensure any standardization in respect of the Guardian system does not cross-over into being anti-competitive.

Potential IP issues

Similarly, if a standard is to arise, Toyota will need to consider whether it is comfortable with its patented technology (behind the Guardian system) being incorporated into that industry standard.  If incorporated, this would render Toyota’s patents standard essential patents, which would in turn, require that competitors be given access (i.e. be granted a licence) of the patented technology on FRAND terms.  Although, what this means will depend upon the circumstances and, to some degree, Toyota’s preference (as a range of different terms may satisfy the FRAND requirements).  There have been some significant disputes between competitors in other industries over licensing of standard essential patents for use in 3G and 4G telecom standards.

Toyota will also need to consider carefully how it wishes to share its technology more generally.  A variety of approaches could be adopted, ranging from selective restrictive licensing to a full open-source model.

The approach taken may ultimately be determined by the existing legal relationships that Toyota has with interested third parties.  For instance, if the Guardian system relies on technology which Toyota has licensed in part from third parties, the terms of those third party licenses may constrain the manner in which Toyota can share the Guardian system.

Potential regulatory issues

Finally it will be interesting to see what regulatory response there is to the emergence of ‘augmented driving’ technology like the Guardian system, and the degree of driving autonomy that new regulations permit.

From a safety perspective, there would appear to be plenty of scope for disagreement on the degree of autonomy that should be permitted, in particular regarding how the safety envelope ought to be defined.  Would this, for instance, prohibit drivers from exceeding the speed limit or from driving when they are tired?  If so, should there be elements of the envelope which can be overruled by drivers?  The latter question has already led to different answers and approaches to fly-by-wire technology in the aviation industry and there is no reason to believe this will be any less contentious in the automotive industry.

Joel Smith
Joel Smith
Partner, London
+44 20 7466 2331
James Allsop
James Allsop
Senior Associate, Tokyo
+ 81 3 5412 5409
Victoria Green
Victoria Green
Associate, Tokyo
+81 3 5412 5432

Lukas Maly
Lukas Maly
Associate, Brussels
+32 2 518 1843

PREPARING AUSTRALIA: CONNECTED AND AUTONOMOUS VEHICLES

In May 2018, Herbert Smith Freehills’ Sydney office hosted the conference Driving forward with connected and autonomous vehicles“, the final in a series of similar events hosted in New York, London and across Asia. The Sydney conference brought together participants from government and the private sector, including infrastructure providers, data analysts, insurers and technology investors, to discuss how government and businesses can work together to best prepare for the advent of CAVs.

Download report

EUROPEAN COMMISSION SETS OUT LEGISLATIVE ROADMAP FOR THE FUTURE OF CONNECTED AND AUTONOMOUS VEHICLES IN THE EU

On 17 May 2018, the European Commission announced a suite of legislative and policy proposals relating to connected and autonomous vehicles. These announcements came as part of the “Third Mobility Package” – the Commission’s broader agenda to ensure “safe, clean and connected mobility” in Europe by 2025 – and are mainly concerned with road safety.

Continue reading

UK GOVERNMENT ACCELERATES TOWARDS A REGULATORY FRAMEWORK FOR AUTONOMOUS VEHICLES

The UK government’s plans to establish a regulatory framework for autonomous vehicles have restarted with the introduction of the Automated and Electric Vehicles Bill to the House of Commons on 19 October 2017. The Bill is the successor to the earlier Vehicle Technology and Aviation Bill, which was scrapped when Parliament was dissolved due to general election earlier this year, and had its second reading on 23 October 2017.

The Bill proposes to extend compulsory car insurance to include autonomous vehicles. Under the draft legislation:

  • primary liability for an accident caused by an autonomous vehicle when in autonomous mode will sit with the insurer of the vehicle;
  • an injured party (which includes the human driver of the vehicle) will be able to claim compensation from the insurer in line with existing practices; and
  • the insurer has a statutory right to claim against any other person who was liable for the accident under existing common law and product liability laws.

By placing primary liability on insurers, the Bill aims to ensure that victims of an accident involving an autonomous vehicle are not left without compensation or forced to pursue complex product liability claims against vehicle and/or software manufacturers. The burden of pursuing such claims will instead fall upon insurers.

The Bill provides two exceptions to this single insurer model:

  1. Insurers and owners of automated vehicles will not be held liable where an accident is wholly due to the insured party’s negligence in allowing the vehicle to drive itself when it was inappropriate to do so.
  2. An insurance policy may exclude or limit an insurer’s liability for damage suffered by an insured person caused as a direct result of software alterations made by the insured person that are prohibited under the policy or failure to install safety-critical software updates.

While the Bill largely replicates the provisions of the earlier Vehicle Technology and Aviation Bill, the requirement that software updates must be “safety critical” represents a tightening of the grounds on which insurers may limit their liability to an insured person.

Another notable change is how autonomous vehicles are defined. As was the case in the previous Bill, the Secretary of State will maintain a list of all motor vehicles that are “designed or adapted to be capable, in at least some circumstances or situations, of safely driving themselves“. However, the definition of a vehicle “driving itself” is now “operating in a mode in which it is not being controlled, and does not need to be monitored, by an individual” – the emphasised wording was not included in the previous definition.

This change goes some way to addressing concerns that were raised regarding the scope of the previous Bill. In particular, it now seems clear that vehicles with level 4 and level 5 autonomous capabilities on the SAE International’s J3016 Standard will be covered, whereas vehicles with level 1 and level 2 capabilities (driving assistance or limited autonomous features) will not. However, the position in relation to level 3 vehicles continues to remain unclear.

Under the J3016 Standard, a level 3 vehicle performs the entire dynamic driving task – including monitoring the driving environment – but the human driver is required to intervene in the dynamic driving task if required by the vehicle. It is debatable whether the requirement for a human driver to intervene if required amounts to the vehicle being ‘controlled’ or ‘monitored’. Absent further clarification, this issue could result in significant satellite litigation.

The Automated and Electric Vehicles Bill is still at an early stage and we could see significant changes before it is ultimately enacted into law. Nonetheless, the Bill marks a significant step towards the UK Government’s aim of establishing a regulatory framework for autonomous vehicles in the UK.

Philip Pfeffer
Philip Pfeffer
Partner, London
+44 20 7466 2660
Anthony Dempster
Anthony Dempster
Partner, London
+44 20 7466 2340
James Allsop
James Allsop
Senior Associate, Tokyo
+81 3 5412 5409