Along for the ride: Considering the legal and practical consequences of self-driving vehicles

In the picturesque suburb of South Perth, a very quiet revolution is taking place. 

If you’re over seven years of age – and have completed an online registration process – you can be part of Australia’s first Automated Vehicle Trial (Trial) , by taking a ride on the Royal Automobile Club of Western Australia (RAC) Intellibus, a fully automated, electric shuttle bus launched on public roads with the support of the WA State Government and the City of South Perth.

The RAC Intellibus travels along a 3.5 kilometre route by the Swan River. This stage of the Trial was launched after initial testing on a private track and then on a public road (albeit without public access.) For this lawyer on board, what is most striking about the ride is how the Trial – and the anticipated future introduction of automated vehicles more generally – will be punctuated by the need to accommodate this new technology within a legal framework.

It already looks different

One of the Trial's aims is to give the public an opportunity to experience automated vehicle technology, and the differences are immediately apparent on entering the shuttle. There is no steering wheel or dashboard, the Intellibus can seat up to 11 passengers and has a maximum speed of 45 km/hour (although the Trial operates at about only 20 km/hour).

Indeed, the shuttle is symmetrical, which enables it to travel either forward or backward with equal ease.

Ordinarily speaking, the Australian Design Rules, which are made under the Motor Vehicle Standards Act 1989 (Cth), set out national standards for vehicles. The Design Rules specify the technical requirements for features such as steering columns, instrument panels and rear vision mirrors – all of which are present to assist human drivers. In the case of the Intellibus these features are either absent or have been modified, so the RAC was required to import the vehicle under the ‘Testing and Evaluation’ category of the Motor Vehicle Standards Regulations 1989 (Cth). This process is different to the discretionary approval of non-standard vehicles, which is also available under the Motor Vehicle Standards Act in some circumstances.

The implication of relying upon an alternative process is that regulatory standards which do not provide for vehicles without a human driver will make it harder to introduce automated vehicle technology. Put another way, if automated vehicle technology is to be more readily embraced in Australia, regulators will need to consider implementing design standards that do not assume human involvement. As technology is continually evolving, regulations will need to be framed to provide clarity and yet accommodate emerging technology. This in turn will require a shift in perspective from automated vehicle technology being seen as experimental to moving into the mainstream.

Just how automated is it?

Automated technology is generally divided into six categories, ranging from no automation (including no power steering) to full automation (which does not even require a steering wheel). The Intellibus is Category 4, ‘High Automation’, meaning that all aspects of driving are automated. The Intellibus nevertheless operates with two Vehicle Chaperones – one of whom holds the vehicle's controller at all times and is able to take control if necessary. (As a side note, and more obvious to the teenagers on board, the ‘controller’ is actually an ‘X-Box’ controller in recognition that some existing technology is already apt for the using.) The Vehicle Chaperones must hold driver's licences and indeed, during our ride, the Vehicle Chaperone had to manually drive the shuttle around an oversized vehicle which was parked outside its designated parking bay.

The need to provide for human intervention reflects the current drafting of the Road Traffic Codes 2009 (WA). For example, rule 263 requires that a person ‘shall not drive a vehicle, unless he or she is in such a position behind the steering wheel that he or she has full control over the vehicle.’ As automated vehicle technology is adopted, the concept of ‘control’ will need to be clarified: is it sufficient ‘control’ for a person to merely be in a position to take back the driving function if necessary? Who then will be considered to be in ‘control’ for the time that that the vehicle is operating in full automation mode? Who is in ‘control’ if the vehicle can be operated remotely?

Just go play in the traffic

It is abundantly clear that the Trial has been set up with safety as the key imperative. The Intellibus employs 3D LIDAR (Light and Detection and Ranging) sensors which use laser beams to measure distance and create a map of its built environment. This allows the Intellibus to interact and react to changing circumstances on the road. It also has 2D LIDAR to detect and avoid moving objects and can automatically apply its brakes if it senses objects in its path.

Anecdotal stories abound of locals familiar with the Intellibus coming within its safety range for the sole purpose of triggering an automatic stop just for laughs. Cheeky locals aside, legislation will need to deal with the apportionment of liability for harm where the technology is literally, the ‘driving force’. It also raises the issue of the extent to which artificial intelligence will be sufficiently advanced to address human ‘cheekiness’: indeed, is there an algorithm for that?

Let’s turn this thing around

The Intellibus operates along a fairly straight stretch of road making a turn at one end. As part of the Trial’s traffic management arrangements, a person stands at the intersection with a sign to stop oncoming traffic, to allow the Intellibus to cross the intersection and make the necessary turn. The contrast between human traffic management with a mere sign guiding such future-forward technology across the road is not lost on those on board.

This traffic management is in place as an extra safety measure to ensure that the automated vehicle does not turn in front of on-coming traffic without human confirmation. The alternative would be for the shuttle to be completely reliant upon its own reading of its surroundings through its navigation and sensory systems before entering traffic. For some, it will take an extraordinary leap of faith in technology to plunge (with family in tow) across an intersection relying entirely on the readings of a machine.

Once safely across, our attention turns to the modifications to our surroundings that may be necessary to allow driverless vehicles to communicate with each other and their surroundings:

• Will sensors need to be placed into all roads, signs and traffic signals?
• What about the inevitable time when traffic is made up of both fully and partially automated vehicles?
• Will I be able to take my automated car to other Australian States with different legislative requirements or will the Commonwealth government have a role to play in road traffic management?
• Australia is made up of millions of kilometres of roads. Will all roads need to be upgraded? Who will bear this cost?
• If the technology relies upon internet access, what will happen in areas where there is no network coverage?

Imagining what could be

One of the best parts of the ride is when passengers are invited to consider ‘what could be’ with driverless technology of the future. We are told that previous passengers have not only included the mere curious, but businesses considering how this technology will affect their operations. Benefits for those who have mobility issues, and increased independence for the elderly quickly come to mind.

But deeper consideration of the potential impact of driverless vehicles requires many assumptions both as to the level of automation, and the way in which it we will adopt it. For example, some argue that driverless vehicles will reduce the need for parking. A driverless vehicle could leave home in the morning, drop its passengers at their various places of work and school, then either return back to its home station or maximise its usefulness by spending the working day contributing to the family budget by providing ride-share facilities for profit.

Taken to the next level, could a driverless car pick you up from work and have dinner waiting for you to eat on the way home (after having self-manoeuvred through a drive-through to collect your order)? Would fast food companies need to modify their sites to accommodate driverless cars? And if this scenario reduces the need for parking spaces, arguably this surplus parking space could be landscaped to allow for greener urban areas. Indeed, it may be that greener streets could address the health impacts of the possible increased congestion due to automated cars driving all day in the ride-share economy rather than parking.

Enjoy the ride

By now this lawyer’s head is bursting with questions and possibilities. With so many ‘what ifs’ surrounding how legislation should accommodate the technology, this passenger could do no more than sit back and enjoy the ride. The telephone calls to colleagues and Government Departments to ask questions and seek clarification will have to wait – for now.

Note: This article was compiled using RACWA Information Sheets ‘RAC Automated Vehicle Trial’ and ‘RAC Intellibus technical specifications.’ Reference was also made to ‘National Guidelines for Automated Vehicle Trials – RAC’s Response to the National Transport Commission’s Discussion Paper February 2017’ and ‘Changing driving laws to support automated vehicles - RAC’s Response to the National Transport Commission’s Discussion Paper December 2017’. The authors also made reference to their personal experience.