Driverless Cars Media Analysis
This media analysis report was written as part of a course at the Centre for Public Awareness of Science. This work represents the accumulative work and skills of a multitude of people and not solely myself.
Section 1: Report overview
Section 2: Executive summary
2.1 — The driverless cars issue in the context
2.2 — Pertinent Scientific Evidence
2.3 — Stakeholders
Section 3: Communicating about driverless cars
3.1 — The driverless cars issue
3.2 — Communicating about risk
3.3 — Driverless cars communication: a focus on Australian media
3.3.1 — Media Analysis
3.3.2 — The campaign — a critical analysis
Section 4: Recommendations
Section 5: Reflection
References
Appendix
Foreword: Our assumptions, attitudes and biases regarding driverless cars
The views and opinions of the authors are reflected in the positive stance towards driverless cars that this report presents, and the assumption is made that a move to driverless cars will be in the interest of society.
Throughout the process of researching this report, the authors gained an understanding of the risks and potential benefits associated with driverless cars. Despite our initial bias, our attempts to keep our analysis even-handed have been sound. We actively sought counter-arguments where information aligned closely with our existing viewpoints. We believe this report to be a fair and accurate representation of the issue of driverless cars and of the campaign in communication.
The authors acknowledge that there was no monetary-induced bias involved in the writing of this report.
Section 1: Report overview
In this report, we examine and discuss risk communication around driverless cars. We focus our attention on automated (autonomous) vehicles (AVs) designed to operate on public roads in the Australian context. Particular regard is given to two aspects of risk communication having relevance in the public domain: the ethics around life-and-death decisions that are to be made autonomously under computer programming; and the economic effects of automation. We consider these the main aspects of interest as this ongoing issue gains momentum.
We have chosen to explore the issue of driverless cars because we see that rapidly-advancing AV technology is intertwined with growing public concern around the safety of AVs and the potential for mass job losses. Recent AV accidents have spiked interest in the current advancements of AV safety, and increased the amount of news articles focusing on the ethical programming of AVs.
Section 2: Executive summary
2.1 — The driverless cars issue in context
- Autonomous vehicles, also known as driverless cars have gone from a demonstrable technology in 2007 to public testing and operation on public roads in the space of less than 10 years. However there are a number of issues with the development and progress of AVs in Australia.
- Australia has historically had a strong identity in car culture, and may be more cautious than most nations in the use and adoption of AVs. AVs have the capability to massively change the face of the transport industry, with far reaching effects such as: making human-based driving jobs obsolete; drastically reducing vehicle deaths and accidents; and changing how we view vehicle ownership − moving from the personal model of vehicle ownership, to shared vehicles.
2.2 — Pertinent Scientific Evidence
- AVs will be interacting with other road users and there is the potential for dangerous situations. In these situations the AV must be able to make life and death decisions about how to act, often this scenario is akin to philosophy “trolley problem” − in essence, the hypothetical situation in which a moral choice is presented between saving one life versus many.
- The issue of job loss due to AVs exists within the greater context of artificial intelligence (AI) automation of labour.
2.3 — Stakeholders
- The rapidly developing AV industry has been evolving and expanding as companies that previously were mainly technology companies are now moving into the forefront of AV development. The sector is becoming increasingly about software and hardware where it has been traditionally considered ‘automotive’. Companies like Google, Apple, Tesla, General Motors and Uber race to be the first to develop a publicly accepted AV system.
- Government will play a key role in the deployment of AV technology with safety regulations needing to be addressed and road rules changes. Car registration and licensing would also be affected.
- With mining companies continuously adopting more automated technology, they will be key financial stakeholders, and also have the potential to help advance the technology faster.
- Driverless vehicles offer Australia’s aging population, and disabled population freedom to travel without relying on others or public transport. For those who cannot use public transport, driverless vehicles offer an easy travel solution.
Section 3: Communicating about driverless cars
3.1 — The driverless cars issue
The concept of driverless cars is not a 21st century issue. Tied with the idea of automation, its roots are deep in history. In relatively recent times, General Motors’ portrayal of a futuristic highway system, the Futurama, showcased the autonomous vehicle to over 5 million people at the 1939 World’s Fair in New York.
But these cars of 1960 and the highways on which they drive will have in them devices which will correct the faults of human beings as drivers. They will prevent the driver from committing errors. They will make it possible for him to proceed at full speed through dense fog (Magic motorways, Norman Bel Geddes, 1940).
For the duration of the 1900s, technology was slow in its development of AVs. This was due to limitations in batteries, computation, hardware and software (Mcarthy, 1970).
The development of AVs picked up speed with the completion of the DARPA Grand Challenge, a competition for AVs funded by the US Defence and Research Projects Agency. Teams vied for a prize by building an autonomous vehicle capable of driving in traffic, performing complex maneuvers such as merging, passing, parking and negotiating intersections. Early on, there were competition years in which no vehicles completed the challenge, and then in 2007, six vehicles finished the obstacle circuit. 2 million dollars was awarded to the top vehicle, and the event was broadcast live to the world, representing a first realistic public presentation of autonomous vehicles in an urban setting (DARPA Grand Challenge, 2007).
With the actualization of AVs through demonstrations and developments in technology, many businesses poured funds into research and development (Lutin, J.M., et al., 2013). In 2009 Google hired members of the DARPA 2007 teams and began development of its own driverless cars in private. In 2013 Google’s driverless car had logged over 50,000 kilometers.
On 16 June 2011, the state of Nevada enacted Assembly Bill №511, instructing the Department of Motor Vehicles to establish the regulations required for the operation of AVs on public roads. These included restricted vehicle operation within specific geographical areas, establishing minimum safety standards and specific testing (Assembly Bill №511). Then in 2012, the first autonomous vehicle license was issued to Google’s Driverless Car in Nevada, followed in that same year by California (DMV, 2012).
There has been a huge number of technology and automotive businesses attempting to secure a position in the AV industry. In 2016 the driverless car branch of google was renamed under the company Waymo, and as of 2018 remains one of the biggest companies at the forefront of AV development.
In the second half of 2016 and the early 2017, there was a flood of announcement about the formation of partnerships, investments, acquisitions and developments. A number of partnerships were formed between AV developers and ride sharing companies, with the aim of offering the first driverless ride sharing service (Navigant Research, 2017). In particular, Uber, one of the largest companies that provides ridesharing services has been developing and testing AVs in the hope of using their current majority share of ridesharing to ‘get there first’.
Tesla is a company known for building electric vehicles that compete with traditional-combustion and hybrid vehicles (Tesla, About, 2018). Most notably, the Tesla Model 3, their first mass production electric car, comes preinstalled with the hardware and software required to be fully autonomous. It offers a semi-autonomous mode called Autopilot. Tesla’s website states “All Tesla vehicles produced in our factory, including Model 3, have the hardware needed for full self-driving capability at a safety level substantially greater than that of a human driver” (Tesla Autopilot, 2018).
On 7 May 2016, a Tesla Model S travelling in autopilot mode crashed into a semitrailer resulting in the death of the driver (Tesla Crash NTSB, 2016). It was deemed in that case that the driver was at fault for ignoring warnings issued by the autopilot system prior to the crash. On 18 March 2018, the first pedestrian collision with an AV occurred, resulting in a large amount of media and public outrage questioning the safety of driverless cars.
The Society of Automotive Engineers released in 2014 a standard on vehicle automation designed to establish a classification system for vehicles that provides levels of automation from no automation to fully driverless. The intention: to help eliminate confusion across disciplines and media; to be consistent with current industry practice; and to help educate the wider community on the role drivers have in performing driving tasks while a driving automation system is in effect (SAE J3016, 2014). (See Appendix 1 for full SAE J2016 Automation Table).
The Navigant Research Leaderboard Grid (Figure 1. 2017 left, 2018 right) ranks many of the companies in the automated vehicle sphere. It ranks companies according to their: go-to market strategy; partners; production strategy; technology; sales, marketing, and distribution; product capability; product quality and reliability; product portfolio; and staying power. Over such a short period of time, many businesses have formed partnerships as AV technology matures, with new companies forming. The number of automated ride-hailing pilot programs is rising as it becomes increasingly clear that mobility as a service will be the primary means of deploying automated vehicles, particularly in the early years of commercialization (Navigant Research Site 2018).
Figure 1. Navigant Research Leaderboard Grid
2017
2018
Cars are very much ingrained in our lives: from general transport to goods freighting; in ride sharing; and in general labour. There are very few areas in our lives that would be unaffected by changes in the automotive industry. Nonetheless this means there is an extremely complex web of business and peoples who either benefit or lose when it comes to complete AVs.
So far we have talked about driverless cars on a broad international scale, not differentiating between locations, although the United States has been one of the prime locations for AV development. Our focus in this report goes to the Australian context, and an introduction to that context begins pre-AV.
Holden and Ford feature strongly in Australia’s automotive history. The first Ford built in Australia rolled off the assembly line in June 1925 (Wikipedia — Ford Australia, 2018). Holden, formally known as General Motors Holden, was founded in 1856 as a saddlery and moved into the automotive industry in 1908 as an importer (Wikipedia — Holden, 2018). In 1948, the company manufactured the first all-Australian motor vehicle (Timeline: Holden’s history in Australia, 2018). On 20 October 2017, Holden ended its Australian manufacturing, but will continue as an importer (Wikipedia — Holden, 2018).
Australia has a strong car culture which is evident in our love of motorsport. From the Australian Grand Prix, run almost continuously since 1928, to the raging battle between Holden and Ford in the iconic Bathurst 1000 (Wikipedia — Bathurst 1000, 2018), Australians have had a deep attachment to motorsport and their cars. Australia is more cautious about AVs than many other nations according to a recent poll (ISPOS, 2018). Ispos director Jessica Elgood says “Perhaps the reluctance of Australians to embrace this emerging technology has to do with our nation having a historically strong identity as a car culture.”
Will car enthusiasts be willing to give up their cars? When musing on that question, we see that consideration of other factors is evoked: there are higher risks associated with a ‘mixed fleet’ (an on-road mix of human-controlled vehicles and autonomous vehicles); and ethical concerns may be higher when there remains human error to consider. Professor Andry Rakotonirainy is the Deputy Director of CARRS-Q and the Intelligent Transport System Human Factor research program within CARRS-Q (“QUT | Staff Profiles | Andry Rakotonirainy”, 2018). He said in an interview,
Theoretically, if we have a driverless car and the driver is out of the loop, it means fewer crashes. In practice, it’s not that easy. […] We face a transition period of a mixture of automated cars, human-driven cars and other road users like pedestrians and cyclists who are not automated. That period worries me (Opray, 2018).
Another influence on the AV issue is our aging population. As people get older and start to experience physical impairments, they can benefit from driverless vehicle technology. A recent study involved thirty elderly drivers ‒ 15 females and 15 males, aged from 70 to 81 years − and analysed the difficulties they experienced during their regular driving. The study identified their needs and expectations regarding Advanced Driving Aid Systems (ADAS) and vehicle automation (Bellet, Paris & Marin-Lamellet, 2018).
The elderly drivers’ acceptance and expectations towards highly automated cars was investigated as a solution to maintaining self-mobility in case of impairments of their cognitive or physical abilities. On average a fully autonomous vehicle had a 50.2% acceptance rate amongst the elderly drivers tested and 38.7% wished to own a fully autonomous vehicle. Although 61.3% of participants did not wish to own a fully autonomous vehicle, 63% said that if, in the future, they did suffer impairments, they would be highly interested in this type of vehicle to keep their mobility (Bellet, Paris & Marin-Lamellet, 2018).
We will return to the discussion of the elderly drivers as part of the following review of stakeholders in the AV issue.
Stakeholders and their influence
Car and technology companies
Car companies and technology companies will be key stakeholders in the driverless vehicle industry. Apple, Tesla, Google have all made advancements in driverless technology with current trials and research underway. In November 2017 Apple published a paper in which they announced a new method for detecting pedestrians and cyclists (Zhou & Tunzel, 2017).
Google has developed the Waymo which stands for a “new way forward in mobility”. Their goal is to make it safe and easy for people and things to move around. Since 2009, their fleet has self-driven more than 5 million miles (approximately 8 million kilometres) mostly on city streets. Waymo is looking to launch a driverless ride-hailing service and also planning to test self-driving trucks (Waymo (Google), 2018).
Companies like Apple and Google, known for advancements in technology, are now able to step into the automotive business, as vehicles become more reliant on programming and electronics.
Tesla has future plans for fully autonomous cars and according to their website, are already preparing their vehicles for full automation when it becomes more widely available.
All Tesla vehicles produced in our factory, including Model 3, have the hardware needed for full self-driving capability at a safety level substantially greater than that of a human driver (Tesla,2018).
Although Tesla is a well-known name in the driverless car industry, Navigant research did not place them in the top 10 leaders of driverless vehicle technology (Navigant Research, 2018). According to Navigant, the top ten leaders in driverless vehicles is as follows:
1. General Motors
2. Waymo
3. Daimler-Bosch
4. Ford
5. Volkswagen Group
6. BMW-Intel-FCA
7. Aptiv
8. Renault-Nissan Alliance
9. Volvo-Autoliv-Ericsson-Zenuity
10. PSA
Not only are these companies key financial stakeholders but the advancement in the technology depends on them. Without companies like GM and Apple progressing the technology, the idea of driverless cars would remain science fiction. They are also vital to the communication of safety as they are at the forefront of safety advancements. They play a key role in communicating new safety measures with consumers.
Mining industry
The mining industry has been adopting driverless vehicle technology in order to improve safety, production and sustainability. Autonomous Solutions, Inc (ASI) is a Utah based company which produces autonomous mining equipment (AZO Mining, 2018). They provide mines with fully autonomous vehicles to work above and below ground. This takes humans out of hazardous jobs and allows them to monitor the production of the mine remotely. Anglo American is a global mining business (Anglo American, 2018) which has entered into a multi-year partnership with ASI to increase safety, productivity and sustainability (Autonomous Solutions Inc, 2018). There are several key stakeholders in the mining industry related to autonomous vehicles. Companies like ASI and Anglo American are key financial beneficiaries. The increase in production and reduced waste that autonomous vehicles can offer mining companies can dramatically increase profit margins while companies like ASI have an opportunity to make large profits on a sort after technology. Additionally, by taking humans out of hazardous mining jobs, there will be less safety risks to workers, reducing death and injury in mines (Chalmers University of Technology, 2018). Mattias Wahde, professor in vehicle engineering and autonomous systems at Chalmers says:
Self-driving vehicles are important in the mining industry for several reasons,[…] one important reason is that you want to minimize the risk of personal injuries by having as few people as possible in the mine. With self-driving vehicles, staff can control and monitor machinery and equipment from a control room” (Chalmers University of Technology, 2018).
This means that the mining industry has high stakes in the AV industry and with potential for mass adoption and purchasing, the mining industry could help speed up the advancements in AV technology.
Aging and disabled population
According to 2013 statistics provided by the Australian government, over a 10-year period, road crash fatalities declined by 24.6 percent, but road crash fatalities in people aged 65 or older increased by 8 percent. Of those deaths, most were drivers or motorcyclists (Bureau of Infrastructure, Transport and Regional Economics (BITRE), 2013). Drivers aged 75 or over, have a higher risk (per distance travelled) of being killed in a crash than any other age group (Older drivers — TAC — Transport Accident Commission, 2018). Driverless vehicles offer many benefits to the aging population and can reduce risk of road crash fatalities. Aiming risk communication at elderly drivers to encourage the adoption of driverless vehicles could result in reduced road crash fatalities in elderly drivers, improved quality of life for our aging population and could speed up the transition period between human controlled vehicles and fully autonomous vehicles.
The aging and disabled population could be key stakeholders in the driverless vehicle industry as they can benefit more than the average person from the technology. Fatal crash involvement by vehicle miles travelled (VMT) starts to increase dramatically by age starting in the mid-60s (Insurance institute for highway safety — Highway loss data institute, 2016). A 2015 survey of Disability, Aging and Caring (SDAC) (Australian Bureau of Statistics, 2016) reported that there were around 3.5 million elderly people in Australia (approximately 15.1% of the population). According to the SDAC, 50.7% of those elderly Australians are living with a disability. Driverless vehicles offer a way for elderly people with impairments to maintain mobility and remain independent for longer (Reimer, 2014). They also offer independence to people with disabilities such as epilepsy or vision impairments. The SDAC reported that in 2015, 40.2% of people living with a disability used public transport (1.6 million people). Of that 1.6 million people, 6.1% could use some forms of public transport but not all and 14.7% could not use any form. The main difficulties that disabled people face when using public transport include: Issues with steps, difficulty getting to the stops and stations, fear and anxiety and a lack of difficulty standing. A report by the NRMA says:
People with epilepsy, narcolepsy, sensory disability, as well as the elderly and the young who are unable to hold a traditional driver’s licence, will be able to use point-to-point transport that meets their needs (NRMA, 2017).
Driverless vehicles offer freedom to travel without relying on others or public transport. For those who cannot use public transport, driverless vehicles offer an easy travel solution.
Governments
A consistent approach will be required for the successful introduction of AVs (House of Representatives Standing Committee on Industry, Innovation, Science and Resources, 2017, p.73). The National Transport Commission noted the importance of working “towards harmonised standards and regulations in relation to AVs to ensure that Australia is well-positioned to adopt new technologies.”
Governments have a role in preparing for the social issues that will eventuate. For example, the current employment model will change. Government, together with industry, will need to provide leadership in order to minimise negative effects of the introduction of AVs. In addition, car registration and licensing will change dramatically with the potential for car ownership to decrease.
We have given a brief history of driverless cars, and begun a discussion of salient issues. Before progressing in the discussion, a consideration of the basis in science and philosophy is timely.
Ethics of AV Programming
Communication of the AV issue is informed by areas including computer science, psychology, economics and philosophy. While technically not science, philosophy is relevant in the understanding of the science.
AVs will be interacting with other road users and there is the potential for dangerous situations. In these situations the AV must be able to make life and death decisions about how to act ‒ even in semi-autonomous vehicles the human driver will probably not be able to take control of the vehicle quickly enough (NSW Parliament, 2016). This means they need a framework for these decisions in order to make the ‘right’ one. Given this requirement for the programming of AVs, ethics plays a big role in the development of and debate about AVs (Faulhaber et al. 2017).
There are two main areas of contention in the discussion and research about ethical algorithms for AVs. The first is which ethical framework should be used in designing the AV algorithms (Gerdes & Thornton, 2016). Although AVs actions will be the result of cumulative decisions by many people and not be moral agents in the same way a human driver is, their actions will still be judged according to the normative moral standards of the society they operate in (Liu, 2017;Gerdes & Thornton, 2016).
Commonly when discussing the ethics of AVs, people refer to a philosophical thought experiment called the trolley problem (Gogoll & Müller, 2017). This is a hypothetical situation in which a moral agent is presented with the choice to save one life versus many, given various constraints such as level of perceived responsibility in taking the lives. In the context of AVs, this could be the choice between swerving to kill one person to avoid many that are directly in front of the vehicle (Gogoll & Müller, 2017).
Many different ethical frameworks have been suggested to guide this kind of decision. These include deontological, Rawlsian, consequentialist and descriptive ethics that mimic human behaviour (Gogoll & Müller, 2017).
Deontological ethics refer to a rule based system, for example one should always do x and never do y. In the context of AVs this could mean always avoiding collisions and never crossing a double line on the road (Powers, 2006). Powers (2006) suggests a model for AVs that follow this system. They suggest various models for implementing this general system and argue deontological methodology is best because of the nature of computer programming is conducive with rule based systems (Powers, 2006).
A Rawlsian ethical system, in the context of AVs, would be one that views the situation assuming the position of each self-interested agent involved and decide on the best possible outcome for everyone involved. This means when deciding what to do it does not ‘know’ which position it occupies in the situation (Leben, 2017). Leben (2017) argues that this method would produce a better survival rate than other options.
Consequentialism states that the most ethical action is decided based on weighing up the costs and benefits of a given action and choosing the one with the best cost:benefit ratio (Gerdes & Thornton, 2016). In the context of AVs this means the vehicle would have a set of weighted values pre-programmed into it, and use these to decide upon actions based on how close to its value an action is. For example, weighing up the cost of taking one human life versus many, or taking a human life versus crossing a double line (Gerdes & Thornton, 2016).
Consequentialism and deontological decision systems are both easily translatable into algorithms and Gerdes & Thornton (2016) examine the practical implications of both, concluding that neither are practical in isolation. Instead they argue for a hybrid system that combines both deontological and consequentialist programming, switching between the two depending on context (Gerdes & Thornton, 2016). Their analysis suggests the decision about ethical systems is more complex than initially presented in the literature.
Given the highly emotional nature of people’s reactions to AVs and the significant role market factors play in the successful implementation of this new technology, decisions about ethical frameworks will need to be made transparently and through engaging with public debate (Australian Parliament, 2017, p.32). One way of achieving this could be with a descriptive ethics approach, which Faulhaber et al. (2017) investigated using a modified trolley problem virtual reality with road vehicles. They found that participants agreed that a consequentialist system would be the most moral and acted accordingly. This is also supported in (Huebner & Hauser 2011). However Bonnefon, Shariff & Rahwan (2016) found that participants would prefer to ride in an AV with a self-interested deontological system that protects passengers at all costs despite agreeing a consequentialist system would be the most moral. This is problematic because if people do not agree with the ethical system in a car they are less likely to buy it (Bonnefon, Shariff & Rahwan, 2016). Thus programming AVs to be consequentialist could reduce their numbers on the roads and fail to improve road safety despite AVs being predicted to be up to 90% safer than human drivers (Gogoll & Müller, 2017). This results in “manufacturers …. [needing to] to accomplish three potentially incompatible objectives: being consistent, not causing public outrage, and not discouraging buyers” (Bonnefon, Shariff & Rahwan, 2016, p.2).
This leads to the second AV ethical issue: whether people ought to choose which ethical framework their AV follows (Gogoll & Müller, 2017). This is the difference between a mandatory ethics setting (MES) or a personal ethics setting (PES) (Gogoll & Müller, 2017). People prefer PES, and are against regulation by government or manufacturing companies (Bonnefon, Shariff & Rahwan, 2016).
It seems that allowing for PES is sensitive to the different moral views people hold (Gogoll & Müller, 2017, p.687). However Gogoll & Müller (2017) argue that not only does this place too much responsibility on the individual, it demonstrates that using the trolley problem to model AV ethics is inadequate (Gogoll & Müller, 2017). This is because the trolley problem doesn’t account for strategic interaction (the fact that the choice is based on social context), iteration (it is not a one off decision, it must be programmed in advance for many situations) or that we could be subjects or objects of targeting (Gogoll & Müller, 2017, p.690). Instead they suggest using game theory as a model, where moral agents are only willing to be moral if other people are being moral, and selfish agents will always choose to value their life above others (Gogoll & Müller, 2017). Under this model, allowing for PES would have a much higher accident rate, and so both moral and selfish agents should want a MES designed to “minimise harm for all people affected” (Gogoll & Müller, 2017, p.695). This conclusion suggests regulation might be necessary but counterproductive to aims of increasing overall road user safety by reducing road deaths (Bonnefon, Shariff & Rahwan, 2016).
Economics of Automation
The issue of job loss due to AVs exists within the greater context of AI automation of labour, and the fourth industrial revolution (Smith & Anderson, 2014). There is considerable disagreement between experts on the potential outcomes of this industrial revolution, where computers are capable of replacing cognitive as well as manual labour (Frey & Osborne, 2013). Thus far there is little systematic evidence on the effects of AI on the workforce and given the research relies on prediction, there is a large margin for error (Williamson et al., 2015). According to Smith & Anderson (2014) half of experts believe approximately 47% of jobs will be lost to automation, and the other half believes the number will be much lower. They also offer this overview of the potential effects of AI on the workforce:
“Key themes: reasons to be hopeful
- Advances in technology may displace certain types of work, but historically they have been a net creator of jobs.
- We will adapt to these changes by inventing entirely new types of work, and by taking advantage of uniquely human capabilities.
- Technology will free us from day-to-day drudgery, and allow us to define our relationship with “work” in a more positive and socially beneficial way.
- Ultimately, we as a society control our own destiny through the choices we make.
Key themes: reasons to be concerned
- Impacts from automation have thus far impacted mostly blue-collar employment; the coming wave of innovation threatens to upend white-collar work as well.
- Certain highly-skilled workers will succeed wildly in this new environment — but far more may be displaced into lower paying service industry jobs at best, or permanent unemployment at worst.
- Our educational system is not adequately preparing us for work of the future, and our political and economic institutions are poorly equipped to handle these hard choices” (Smith & Anderson, 2014, p.1–2).
It is important to note that the fear of technology replacing jobs is not new, and automation has always been met with resistance from various self-interested parties (Frey & Osborne, 2013, p.5). According to Mokyr, Vickers & Ziebarth (2015) there are three common forms of technology anxiety:
- Replacement of labour leading to widespread unemployment
- Widespread unemployment could be bad for the human psyche (although this is most likely a myth designed by the ruling elite to maintain inequalities)
- The “epoch of major development is behind us” (Mokyr, Vickers & Ziebarth, 2015, p.32).
In regards to AVs in particular, the “world economic forum estimates AVs will displace 5.1 million jobs across 15 major economies by 2020”(Australian Parliament, 2017, p.47). Furthermore, CEDA predicts AVs will replace up to 5 million Australian jobs in the next two decades, either directly or indirectly (Australian Parliament, 2017). Those most likely to be directly affected by AVs in the workforce are professional drivers. In 2015 approximately 247,000 Australians were employed driving trucks, buses and taxis (Australian Parliament, 2017, p.49). Whilst taxi drivers will be the most affected they will likely still be necessary for customer service roles, such as assisting the elderly and disabled (Australian Parliament, 2017). Another mitigating factor is that given the shift will likely take decades it will probably be accompanied by more jobs being created in other industries. Furthermore, instead of displacing current drivers, the move away from professional driving may simply reduce the number of people entering the profession, allowing current drivers to keep their jobs until retirement (Australian Parliament, 2017).
The Australian Driverless Vehicle Initiative (ADVI) predicts that AVs will also produce many economic multipliers, and what these are depends upon the exact ways in which the current supply chain for vehicles changes (Haratsis, 2016, p.4). According to ADVI there are two models for the introduction of AVs: high and low disruption. They predict the high disruption model, with AVs making up 15% of new cars sold by 2025, is more likely (Haratsis, 2016). On this basis they predict Australia will generate $15 billion in revenue and generate 7500 direct jobs and 16,000 indirect jobs from AVs (Haratsis, 2016). Based on this data I conclude that whilst experts disagree about the extent of the impact AVs will have on job loss, the impact will likely be significant. However this is likely to be a long term effect, and not necessarily bring to fruition the fears discussed above.
We have now the background to the AV issue. We move to a review of theory best suited to an understanding and analysis of the risk communication of the issue in the public domain.
3.2 — Communicating about risk
With the prospect of AV technology greatly improving road safety, how should the overall risk be communicated in the public realm? Human error is currently the main cause of road accidents (House of Representatives Standing Committee on Industry, Innovation, Science and Resources, 2017, p.24). It is anticipated that AVs will greatly reduce accidents by removing the requirement for real time decisions by humans, instead placing those decisions with the computer algorithm.
Going driverless — ethics in AV programming
We have introduced to our discussion the dilemma known as the trolley problem (Gold, Colman & Pulford, 2014, p.1), which must be dealt with in the programming of AVs. One question being asked is whether there should be a mandatory ethics setting that all AVs are set to, or whether each ‘driver’ should have the choice to select a setting. Communicating about considerations like this will require a set of tools.
Under the mental models approach, ‘risk communication’ means communication intended to supply laypeople with the information needed to make informed, independent judgments about risks to health, safety and the environment (Morgan, 2002). The implication for those who communicate risks is that effective communication requires an understanding of what the audience already believes about the risk (Lundgren & McMakin, 1998: p.17). The communicator identifies the audience(s), researches their views on the risk, and compiles the results to build a mental model for comparison with ‘expert model’. The idea is that a useful risk communication is developed in this way — evaluation is critical.
According to Miah (2005), […] the salient aspects of science and technology have less to do with technical details than with ethical implications…the public(s) are more concerned with what science ‘means’ than with how it ‘works’. He asserts that ethicists, rather than scientists, should do the communicating. Lamberts (2012 p.146) calls for pragmatism in applying ethics to real world situations…if becoming mired in theoretical debate, there will be no action.
From this basis the public perception of the risk can be shaped. The concerns of the public will need to be addressed if the benefits of AVs are to be realised. There is arguably an imperative, born of a strong potential to save lives, to avoid delay in the move to AVs. The success of communication efforts should be monitored and evaluated against agreed goals and their measures, with results used to inform the ongoing efforts.
In developing our approach to the communication of this risk in the public domain, we note the differentiation of that domain by Sandman (2003) into ‘publics’ — people who don’t care much about the issue — and ‘stakeholders’ — people who have a stake in the issue, and who know they do. Sandman points out, however, that those who have a stake but don’t know it are probably deserving of stakeholder status. I suggest that for the risk discussed there will be a large cohort of publics in that latter category. We will have the attention of stakeholders cognisant, but we should engage early the less self-aware.
According to Sandman (2003) one can distinguish four types of public involvement: fanatics; attentives; browsers; and incentives. Under that breakdown the inattentives are those who don’t care, and can be largely ignored. He notes though that as the issue becomes more prominent, uninvolved neutrals can become suddenly attentive and sceptical. It is preferable that they are engaged early and receive a balanced presentation. “The browsers and the media are a matched set” — periodic coverage from television and newspaper will satisfy them. The situation is similar with the attentives, but with supplements from specialty media. Fanatics should receive the focus of attentions.
Sandman, neatly summarising, asserts that if a stakeholder seeks direct contact, they should receive it. “Interact with the fanatics while the attentives watch. The browsers follow casually in the media, and the inattentives don’t know it’s happening” (Sandman, 2003, p.7).
For most Australians, vehicles that operate in a mode requiring no human monitoring or intervention are an unfamiliar technology. “[It is likely that] to achieve the required level of social acceptance, reliability needs to be demonstrated through pilots and public participation” (submissions to House of Representatives Standing Committee on Industry, Innovation, Science and Resources, 2017, p.21). The Standing Committee has recommended AV trials on public roads. They note that the ‘trolley problem’ scenario is being highlighted in the media and needs some attention in order for the general public to feel sufficiently safe that they are willing to ‘grant’ the technology a ‘social licence’ to operate.
The notion of availability may apply to our understanding of how people assess risk in everyday life. Through the availability heuristic we assess the likelihood in risk by asking how readily examples come to mind (Thaler & Sunstein, 2008). The level of concern, the estimation of the seriousness of a risk, will likely be higher if examples come readily to mind. With recent mass media reporting of AV accidents, and in the absence of sound risk communication, ‘availability bias’ can be at play.
Considering again the ‘perception’ of this risk, we note that the severity of the actual hazard is high, manifest in road accidents, but that likelihood is low. Hence, by combination, the risk is low. A perception of a relatively high vulnerability in riding in a small vehicle, conspicuously without driver, would arguably explain an increased level of stakeholder arousal (Sandman, 2003), and increased perceived risk. One might consider whether this specific perception yet exists in the public arena.
Control of cultural cognition may be achieved through having risk communication vouched for by a diverse set of experts. This is because it feels safe(r) to consider evidence with an open mind when a knowledgeable member of our own cultural community accepts it (Kahan, 2010).
An additional outcome of such control may be to confirm, or correct an existing judgement arrived at by the expert heuristic. Further, I suggest that this may provide a countering of the cognitive bias around the likelihood of the risk scenario — although media reported AV accidents may not have been of the trolley problem type, they may in that case have been conflated with that scenario, thereby contributing to a bias there. The control method described may act in that also.
In operation together with these factors, a key contention of the social amplification framework is that signals about risk may be modulated in passing through social ‘amplification stations’ (scientists, the mass media, government agencies and politicians, interest groups), resulting in intensification or attenuation of risk in ways predictable from social structure and context (Pigeon, 2014, p.353). Careful framing of our risk communications may be a powerful tool of modulation. Even experts are subject to framing effects (Thaler & Sunstein, 2008).
Regardless of the content of the risk communication, ‘intelligence’ in the use of language warrants a generous commitment of resources. In particular I think of the persuasive power of repetition (Romm, 2012). Another aspect of language use that may have applicability in our context relates to the avoidance of the perpetuating of myths around autonomous entities generally. If wanting to debunk a myth, a focus on truth is important, instead of repeating the myth (Romm, 2012).
Relating the risk around employment
As we have discussed, there is agreement that the introduction of AVs will have a significant impact on employment. There is disagreement as to the extent, but most agree there will be at least some job losses and changes in employment patterns. It is anticipated also that new employment opportunities will arise (House of Representatives Standing Committee on Industry, Innovation, Science and Resources, 2017).
The development of risk communication in this will draw in part on theory already discussed. Theory considered in the following will further our understanding and analysis of the communication effort.
People tend to resist scientific evidence that could lead to restrictions on activities valued by their group, but if they are presented with information in a way that fits with those values, they react more open-mindedly. Group-specific risk communication should be presented in a manner that affirms rather than threatens values, and with careful consideration to salience (Kahan, 2010).
The danger of job loss and/or job reassignment is real and will affect people to varying degree. Loss of any kind can be challenging, and people’s automatic systems can be thrown into turmoil. People are ‘loss averse’, loathed to make change, even change that is in their interest (Thaler & Sunstein, 2008). Self Characterisation Theory (SCT) tells us that as humans we characterise elements of our ‘life experience’ (objects, emotions, people) to make sense of our world and to make comparisons. We identify with groups we see ourselves belonging to.
A need for crisis communication (High-Hazard, High-Outrage) may well occur in this aspect of the issue. People are likely to be very upset — they are appropriately upset stakeholders and may need help bearing and expressing their emotions. Acknowledgment of their uncertainty may help (Sandman, 2003).
This review of theory relevant to the issue of the driverless car has been with particular regard to the ethics around life-and-death decisions made in AV programming and to the economics of automation. The theory will be applied in sections to follow.
3.3 — Driverless cars communication: a focus on Australian media
3.3.1 Media Analysis
The aim of this media analysis is to gain insight into how risk in AV-related safety and job automation is being communicated to the Australian audience. Our media focus is particularly concerned with messages that reach the larger proportion of population, and whether those messages are skewed (positively or negatively).
Search Terms and Restrictions
Our search terms for data gathering were: driverless cars; autonomous vehicles; and jobs or ethics. Of all Australian households, 86% are connected to the internet (8.1 million, 2016), with over 90% of Australian internet users using Google as their primary search engine(StatCounter, 2018).
Due to the high Google use by the Australian population our terms were chosen by using Google Trends — a tool that shows how often a search term is used relative to the total search volume; We restricted the tool to searches from within Australia, on the topic of driverless cars, and for the period of 2008 to present. The timeline restriction is because of limits on Google search data on AV as a topic.
Google Trends (Figure 2) shows the most googled words around the topic of AVs are autonomous, driverless cars, and self driving cars. By using similar terminology and keywords, we can better align the media we find to that which the populace would discover.
Stakeholder Impact and Media/Public Attention
Further analysis of Google Trend data (Figure 3) shows AV-related searches beginning to rise in March 2014 and spiking in May 2018. The highest spike in searches coincides with the first pedestrian death by AV (Uber), with the main search terms during that period being driverless car death, uber self driving car and driverless car. At this time, searches on the topic of driverless cars hit a peak high.
Figure 3: AV google search trends over time (Google Trends ).
Looking more closely at both the most common search terms and the increase in searches on related topics over time, there is a strong media presence of a few main stakeholders, in particular Tesla, Uber and Google (Figure 4).
Choosing Media Sources
Australians consume news at variety of sources, mainly through television, social media and websites/apps. Our media was chosen from the top 5 most used media platforms news.com.au, ABC News Online, nine.com.au, Yahoo! 7 and the Sydney Morning Herald (W. Jerry Et Al, 2016).
News organisations are becoming increasingly cross platform, with articles and stories relayed through multiple social media channels, digital newspapers, TV and radio. Likewise, consumers are using social networks more than ever to get news media, with a significant portion of consumption through Facebook and Youtube (W. Jerry Et Al, 2016).
According to Facebook statistics by Socialbakers, ABC News is the top Australian media page on Facebook, garnering a fan base of around 3.5 Million. In addition to the media organisations, 6 of the top viewed videos from Youtube connected with AV search terms had garnered a total of over 25 million views.
It is important that our searches for media occur directly at the organisation's repository, in order to avoid the ‘echo chambers’ and ‘search bubbles’ of Facebook and Google’s search engine.
Codifying , Categorization and Analysis
Our media selection was collated into a spreadsheet table. The data gathering consisted of two parts, the first being more quantitative, broad article data, and the second being more qualitative message/media characteristic data.
Broad Data
- Section — differentiates whether an article’s risk issue focus is on the issue of death ethics or on Job Automation, this is determined by whether the content discusses automation of jobs or death associated with a AV, in the case that both are talked about then it adopts both sections.
- Date — The day, month and year the article was published
- Position — Position is a measure of an articles stand on the issue, with the position spectrum divided into Against, Cautious Suspicion, Neutral, Cautious Optimistic, and Supportive.
Section
Out of the articles reviewed an overwhelming number (84%) of articles focused on the ethical side of AV risk issue in contrast with just 16% of articles talking about the risks on Job Automation. This shows the safety and ethics having far more of a media presence and immediate outrage factor.
Date
From the google trends data, we can expect the volume of articles to increase as awareness and major events of AV risk move closer into the public’s awareness, this is demonstrated by a rising increase of articles from 2014–2018, with 64% of articles occurring in 2017 and 2018 a key time period for major AV events like the Uber and Tesla crashes.
The Position Spectrum
Supportive
Statements that suggested a positive or advantageous position, swayed the position metre further towards supportive — statistics or statements pointing to AVs improving on aspects such as crash rates, vehicle emissions, or providing access to people in need.
- AVs “Significantly reduce crashes and congestion and also reduce vehicle emissions and fuel use.” (Queensland is preparing for driverless cars to hit the streets).
- “A primary argument for driverless vehicles is safety.” (The №1 reason we should put driverless cars on the road).
Neutral/Cautious
Neutral articles often pose questions or are mainly reporting on research without any obvious sway towards either side of the position spectrum.
- “The trial is set to continue, but the incident will do little to ease the concerns of those sceptical about self-driving cars…. It shows that we need to have this discussion, and have it quickly.” (Self-driving bus crashes in first hour of service).
- “But is this the car’s fault, or the programmer’s? Or is it ours?” (Driverless cars: We don’t know what we want when it comes to accidents).
Against
Words and statements with the intention to strike fear, or link aspects to strong negative events like terrorism or explosives, put an article well towards against. Similarly, if a sentence makes a statement about driverless cars and concludes with a negative impact or affect.
- “AUTOMATIC self-driving cars, which will be seen on the streets in Melbourne in the coming decades, could be sabotaged and used as weapons by terrorists….Police fear terrorists could get a driverless car, pack it with explosives,….” (Why driverless cars could become a terror threat to Melbourne).
- “IT’S one of the most exciting technological advances on the horizon, but the arrival of driverless vehicles could devastate Australia.” (The jobs killer is coming: How driverless trucks could change Australia).
Figure 5: Article Position Distribution
Our position distribution is shown in Figure 5, with 36% of articles taking a supportive stance, welcoming the technology — benefits far outweighing the risks. 48% were optimistic but cautious, acknowledging the possible benefits but reluctant to embrace or adopt a fully supportive stance. 8% were neutral, reporting on research having no definitive sway towards one way or the other. 4% were cautiously suspicious, being concerned with emphasising negative issues. Lastly, 4% were against the topic and rejected the technology. Whether articles fall into one of the five categories as indicated by the majority presence of positive words like excitement, amazing and welcoming or negative emotional words such as terror and concerning as well the logical argument structure that situates a message along our position spectrum.
Characteristics of Messages and Media
The focus of our media breakdown is on messages, messengers, evidence use and media characteristics. In analysing multiple media from 5 media organisations, we anticipated seeing differences in the tailoring of articles to suit audience and readership demographic:
- On March 18 of this year, an Uber driverless vehicle hit and killed a 49 year old woman who was walking across the street with a bicycle. This incident resulted in an increased number of news articles focusing on the safety of AVs and the ethical concerns around AV programming.
Are there one or more messages? What are they?
- Safety was a main message overall when it comes to ethical concerns, often accompanying crash statistics, and jointly expressing that safety is also a concern. The message that “driverless cars could be safer than traditional cars, but that safety is an inherent concern” make for a somewhat contradictory argument.
- “Driverless cars are safer than human drivers and will reduce road toll” (The №1 reason we should put driverless cars on the road).
- “Autonomous vehicles are coming in the future but safety is still a concern.” (One million driverless vehicles by 2035: NRMA).
- “People would be reluctant to buy a self-driving car programmed to let them die in order to save pedestrians, according to a new study…. Ethical question is still an issue“
- The messages surrounding job automation where few but inherently negative emphasising disaster and personal loss.
Messenger’s intentions? (explicitly/implicitly)
- Positive messages here generally have the intention of reassuring the public that AVs are safer than human drivers.
- Negative messages have the opposite intention of expressing how unsafe AVs can be, and also raise the concerns over ethical programming.
- Researchers giving statements on their research
“Communicate results of the authors research and bring more public attention to the research results and ethics issue” - The intention of some articles on major events such as the Uber pedestrian crash appear to be to reduce public outrage, minimise negative perceptions of AVs and reduce reputation damage to AV development stakeholders.
Specific messengers?
- Messengers include car and tech companies, roads ministers/government officials and companies trialling AV systems.
- Transport Workers’ Union, Honda Manager, UTS students, Ian Christensen, the head of iMove, Volvo Australia’s technical manager, David Pickett, NSW Roads Minister Melinda Pavey, Associate professor Hussein Dia from the Smart Cities Research Institute at Swinburne University of Technology Dr Zubair Baig, a senior lecturer in cyber security at Edith Cowan University
Evidence use
- Sources of evidence often include academics (Doctor, professors, graduates), business people (research companies), managers and CEOs of stakeholders companies, statistics and studies as well as video footage.
- “Not all companies are ready,” said associate professor Hussein Dia from the Smart Cities Research Institute at Swinburne University of Technology. (Driverless cars are now tested on our roads, so should we be worried?)
- “This futuristic vision is not a matter of “if”, but “when”, according to RAC senior manager policy and research Anne Still” (Is Western Australia ready for driverless cars?)
- “Honda’s General Manager for customer and communication Scott McGregor”(Technology behind the collision-free cars of the future)
Restrictions and limitations
There was a variety of restrictions and limitations on our gathering of data these included the use of news websites to avoid the search bubble effect of google search engines. As currently most AV development and major AV events have occurred internationally, outside of Australia, our analysis has largely omitted international messages, this has been due to limited resources.
The availability of data is also a limiting factor, the distribution of media on AV issue is present in a large variety of news organisations and media types (radio, youtube, facebook pages, twitter ect), data on readerships, view count and geographical locations is often private and not publicly accessible.
Evaluation — has there been any?
In our review of relevant theory, we discussed the application of mental models to the communication of risk. Under that approach, the process of developing useful risk communication of the AV issue involves the construction of an ‘audience model’ for comparison with the ‘expert model’. Evaluation is critical to that development, and will ideally be iterative. Around the AV issue, we have not found evidence of direct evaluation of messages by their communicators. However, there are articles that review the evaluatory discussion of communication needs.
A recent article points to a survey in which 78% responded that they feared the prospect of riding in a driverless car. A poll by a large US insurance firm found that 41% were averse to sharing the road with AVs. They noted that surveys by MIT and others showed that “ironically, even as companies roll out more capable semi-AVs, the public is becoming less — not more — trusting of AVs” (Hutson, 2017). Intel’s chief AV architect lamented that they could offer the safest car in the world, but that if consumers won’t put their families into it, there is no market. Consumer distrust has sparked a wide range of studies into people’s perceptions of AVs — ”and what might persuade skeptics to change their views.”
Celebrity basketballer LeBron James recently featured in an AV advertisement aimed at overcoming consumer distrust of the technology. Known for being unshakable on court, James at first declines an AV ride, but after a short trip exclaims “Hey yo, I’m keepin’ this!”
Hutson cites an Intel study finding that “familiarity will ease some anxiety.” It was found that if a car is given personality, in the ways that robots often are, the car ‘invites’ trust.
Another study that was identified as evaluating the shortcomings in communications about AVs, recorded the responses of bystanders to a “ghost driven” car. The team at Stanford (Rothenbücher, Li, Mok, Sirkin, Ju, 2016) looked at the effect of bystander’s trust in AVs on the take-up of the technology by consumers. “Nobody wants to be the [jerk] driver, even if they’re not actually driving the car,” says Ju. In the ghost driver system a non-AV car is used, and the driver wears a special seat cover to make it appear as if there is no driver. One finding was that, at pedestrian crossings, pedestrians want some feedback from an AV that it has ‘noticed’ them. If they don’t receive cues like that, they will even go out of their way to avoid the car.
After becoming aware of the ghost driver studies, Ford motor company broadened their research focus to consider driver and customers, but also how society in general will interact with their vehicles (Hutson, 2017).
If we consider then the trolley problem, Huston notes that according to a researcher at MIT,
When asked about their top concerns about AVs,[…], most people don’t mention trolley problems. And even telling people explicitly about the dilemma doesn’t necessarily enhance their fear of AVs or reduce their desire to buy one ( Rahwan, 2017).
Still, media coverage could yet shape public opinion. One prestige car company “faced public indignation” in response to an official’s comment “If you know you can save at least one person, at least save that one. Save the one in the car.” the company admitted that AVs would run over a child rather than putting a passenger at risk by swerving. The company was quick to back pedal, but we are reminded that “public outrage is difficult to predict. […] resistance to non-utilitarian cars could end up being a big deal.”
Trust in AVs may come through giving a balanced account of the technology. The term ‘calibrated trust’ is now being used in AV literature. Trust calibration is the process of balancing user trust to the required level (Mirnig, A., Wintersberger, P., Sutter, C. & Zeigler, J. 2016, p.34). “You want people to trust the AV […] with respect to the things it’s actually good at, but not trust the AV with things it is not good at.”
A recent poll shows that one in six Australians would never use an AV, and findings suggest that “a wait-and-see approach” is being taken (Wade, M., 2018). However, a survey across 28 countries (n = 21,000) found that Australians are more open to the possibilities of AV technology than the US, UK, Germany and Japan (Ipsos, 2018).
It does appear that the evaluating of communication efforts on the AV issue is increasingly viewed as important. We go on now to critique the campaign to date.
3.3.2 The campaign — a critical analysis
The campaign around AVs is a complex mix of stakeholder views and strategy, but it has taken a clear direction in favour of adoption. We now consider how effective the campaign has been in communicating risk in the public domain.
We analyse the top 5 news sites and critique the communication efforts surrounding the ethical concerns of job automation and the safety of AVs. We have split both categories into positive, neutral/cautious and negative, with a critique of the communication for each of those positions. To accurately critique the communication efforts we have focused on three main concepts, use of evidence and credibility of a speaker or source (Ethos), emotionally driven arguments (Pathos) and logic-based arguments (Logos). We draw on Sections 3.1 and 3.2 of this report.
Positive safety messages
Positive safety messages tend to use logical arguments and facts to show that AVs are safer than human drivers and they use strong emotional language when describing facts such as road death tolls. In the Sydney Morning Herald article “The №1 reason we should put driverless cars on the road” they use statistical facts with strong emotional language, for example “Consider the current situation. In 2017, 1225 people needlessly lost their lives on Australia’s roads”. Here they have tried to evoke emotion by saying “needlessly lost their lives” instead of just stating the statistic. The article is written by Dr Stephen Greaves professor and chair of transport management at the University of Sydney Business School’s Institute of Transport and Logistics Studies which adds credibility to their article. The intention of an article like this is to convince readers that AVs are safer than human drivers and can reduce road death tolls significantly.
Another example of the Sydney Morning Herald using this style of communication can be seen in their article “Queensland is preparing for driverless cars to hit the streets”. In this article they quote Main Roads Minister Mark Bailey “Automated vehicles can and will I believe reduce the risk of crashes linked to alcohol, distraction, drugs and fatigue, making our roads much safer,”. Mark Bailey’s title gives him credibility on the topic while his words evoke emotion and focus on the human errors that cause car accidents.
Neutral/Cautious safety messages
Neutral articles explain both the public’s concerns and the reassurance of safety expressed by experts. Their intent is to encourage discussion about particular issues. They write mostly in the ethos style, using direct quotes and statistics. The ABC are the most neutral on the issue. They provide facts and direct quotes, and tend to finish their articles in a way that leaves the reader wanting to discuss the topic further. They do however use some strong emotional language in their titles, to draw attention. One example is the article “To kill or not to kill: Researchers probe moral code for driverless cars as technology hits Australian roads.” (Donoughue, 2015), they’ve used some strong wording in the title to draw attention, but the article goes on to discuss a recent study into the public’s acceptance of a utilitarian AV. They use direct quotes from the study’s author and present a very fact based article. It is unbiased and leaves the reader open to discussion about the ethical programming of AVs. Control of cultural cognition is achieved — it feels safer to consider evidence with an open mind when a knowledgeable member of our own cultural community accepts it (Section 3.2, p.XX)
In another example, the ABC opens with a short and punchy intro about an incident and its possible effects, to draw in readers. “An Uber self-driving car has hit and killed a woman crossing the street in Arizona, marking the first time a self-driving car has killed a pedestrian and dealing a potential blow to technology which is expected to transform transportation.” (ABC, 2018). This article, “Uber suspends self-driving car tests after vehicle hits and kills woman crossing the street in Arizona”, states with pathos “Local television footage of the scene showed a crumpled bike and a Volvo XC90 SUV with a damaged front.” A persuasive image of the scene evokes emotion.
Also in the article, “The cars can react much faster than human drivers to dodge potentially dangerous situations and have the ability to predict when a collision may occur” — a logical argument that AVs can react faster than humans and predict collisions. Later we see, “Concerns over the safety of autonomous vehicles flared in July 2016 after a man was killed while travelling in a Tesla partially self-driving car which collided with a truck in Florida. Safety regulators later determined Tesla was not at fault.” Here they’ve used logos by stating the fact that Tesla was not at fault. This implies human error again. The intention of this article is to insight discussion around the safety of AVs.
Negative safety messages
In an article by news.com.au “Why driverless cars could become a terror threat to Melbourne” (Hurley, 2016), they open with the line “AUTOMATIC self-driving cars, which will be seen on the streets in Melbourne in the coming decades, could be sabotaged and used as weapon by terrorists”, employs strong language, including the emotionally charged phrase ‘weapon by terrorists’. Coupling those words with ‘self-driving cars’ incites fear, drawing the reader in. The capitalising of ‘AUTOMATIC’ evokes the ‘rogue robot’ myth from popular culture, and the word happens to be positioned diagonally adjacent the word ‘weapon’ on the page.
Adding credibility to this claim, “Australian Federal Police Deputy Commissioner Ramzi Jabbour warned smart cars being developed by companies such as Google could be exploited by criminals, including extremists, to wreak havoc.” Readers are then given a scenario they can picture, “Police fear terrorists could get a driverless car, pack it with explosives, pre-program it, then days later from the other side of the world use a computer to activate the vehicle along a course to a specific target.” Juxtaposed confusingly with this message, we are told the “Victoria Police Chief Commissioner Graham Ashton said he believed the advent of the self-driving car would ultimately save lives.” Conflicting messaging in the communication makes the intentions unclear. Very strong wording draws readers in but leaves you confused about whether AVs are potentially safe or unsafe.
Negative job automation messages
The article “The jobs killer is coming: How driverless trucks could change Australia” posted on new.com.au (Reynolds, 2016), clearly intends to worry people into reading on. They maintain that hard-hitting message in stating that “IT’S one of the most exciting technological advances on the horizon, but the arrival of driverless vehicles could devastate Australia.”
People are ‘loss averse’, our automatic systems can be thrown into turmoil (Section 3.2, p.XX). “As a nation whose economy revolves around a $200 billion transport and logistics industry, waving goodbye to truck drivers could mean far-reaching effects.” This line is intended to also send fear into the wider community. “It’s not just drivers who will be out of jobs. Businesses linked with truck drivers and the roadside will struggle or close too, from service stations to petrol stations to cafes to moteles.” People tend to resist evidence that could lead to restrictions on activities valued by their group (Section 3.2, p.XX). Quotes from the Transport Workers’ Union are used to add credibility to their fear-raising claims.
Neutral/cautious job automation messages
Neutral communication about job automation on the top 5 news sites is minimal. ABC are once again the most neutral on the topic. In their article “Driverless cars to bring job losses, hacking worries and ethical questions — but get used to them” (Lowrey, 2017), they start with a very strongly worded title to draw attention, but then the article discusses both the positives and negatives of AVs and job automation. Direct quotes from the Standing Committee on Industry, Innovation, Science and Resources chair Michelle Landry, add credibility to their article. They use subtitles such as “Jobs to come, but more to go” to keep the reader interested, but end on a neutral note — “Ms Landry puts forward an optimistic view, that it is hard to predict the jobs of the future”. They point to it taking decades for the technology to fully roll out, and that many people currently in driving occupations will have retired by the time their job becomes redundant.
Positive job automation messages
Positive job automation messages on the top 5 news sites that we analysed are non existent. The American Yahoo news site has several positive articles around job automation such as “Driverless cars will make a lot of jobs better, not destroy them” (Quartz, 2017) but the focus of this report is on Australian news sites and the communication to the Australian public, which does not directly target foreign news sites. Australian drivers and mine workers are at high risk of job loss with the implementation of AVs, and we discuss in our recommendation section ways to improve positive communication efforts surrounding job automation in Australia.
Does it work? — people’s views on AVs, intentions vs perception
Australia’s acceptance of AV technology is behind that of other countries (Ipsos 2018). This could mean that positive messages about the issues are not as effective as the negative messaging. Positive messages seem to focus in a logos and/or ethos style in which logic and facts are used to convince the reader that AVs are safe. This may not be the best approach, as it seems many people still have concerns around the safety and ethics of AVs.
Negative messaging uses a combination of ethos and pathos, generally using strong emotional language and emotional stories (such as the ‘Uber’ death) to evoke emotion and fear around AVs. They sometimes use quotes from government officials, car companies and videos of accidents to support their articles. News.com.au were the most negative on AV safety, with strongly worded articles (p. XX) that focus on evoking emotion. By using direct quotes they are able to give readers a sense of confidence in the accuracy of the article. With growing concerns over the safety of AVs, this article draws in a large reader base and can have a negative impact on the public’s acceptance of AVs. We discuss in our recommendation section how positive safety messaging could adapt a more emotional style of communication and target different social groups in order to have a larger impact on Australia’s AV acceptance.
As noted earlier in this section (p. XX), we have found little evaluation of message effectiveness by communicators. Assuming this is true, it is a major flaw in the campaign, and is discussed further in Section 4.
Section 4: Recommendations
In this section we offer recommendations to improve risk communication efforts and reach a wider audience, with the intention of increasing Australia’s acceptance of AVs.
Recommendation 1
The success of communication efforts should be monitored and evaluated against agreed goals and their measures, with results used to inform the ongoing efforts.
Recommendation 2
Given the strong emotional attachment that Australians have with their own vehicle, there will need to be acknowledgement of that attachment in any plan. It may be useful to draw from an analogous historical example — people still enjoy riding horses, but have successfully transitioned from riding on the highway. Greater accessibility to motoracing parks may be attractive, and should be explored. People are ‘loss averse’, and communications will need to acknowledge this. Group-specific risk communication should be developed, affirming rather than threatening values.
Recommendation 3
That the Commonwealth Government facilitate/encourage AV trials, with a particular focus on trials that enable a public experience of AVs on public roads. That governments Australia-wide fund trials of AVs with a public transport application. Government needs to take the lead on promoting outreach and acceptance — the real world experience of seeing these trials will have people wanting to get on board. The message needs to be more immediate, tangible, as against futuristic/distant.
Recommendation 4
Given the highly emotional nature of people’s reactions to AVs and the significant role market factors play in the successful implementation of this new technology, decisions about ethical frameworks will need to be made transparently and through engaging with public debate. Engaging experts , for example universities, in trials and discussion around the ’trolley problem’, to help in developing familiarity. There may be an opening for the careful use of satire — “AVs save 100 lives in wake of 300 road deaths this morning.”
Recommendation 5
Following from Recommendation 2, enthusiasts’ retention of their traditional cars will have a bearing on the problem of mixed fleet. What strategies can we use to communicate the risks of mixed fleet? Robotics professor Jonathan Roberts, says that the best approach during a transition period would be for the government to give 5 years notice that all cars on the road must be automated (Opray, 2018). It will be important to consider this type of approach when convincing Australians to give up their cars for AVs. Risk communications should be carefully framed to attenuate risk perception.
Recommendation 6
Increase communication around job automation and the opportunities that the AV industry can provide. Provide clear communication about job opportunities for those currently in driving jobs. The Government should implement training opportunities to ease driver’s transition into the AV industry and reduce job losses. A need for crisis communication (High-Hazard, High-Outrage) may well occur in this aspect of the issue.
Recommendation 7
Increase our aging population’s acceptance of AVs, with more targeted communication efforts. Additionally, similar communication efforts can be targeted towards Australia’s disabled population as they too can experience significant benefits from AVs. In line with these efforts, that the Commonwealth Government’s preparation integrate the needs of these people. These particular communities would benefit from a more emotional style of communication in which the personal benefits to them and their families is made clear in a compassionate, caring way. Strong, down the line facts may not be as effective.
Recommendation 8
That the campaign, in part, focus towards showing people how much more control they have over their daily life with AVs: more time to get things done; less money on parking; and more freedom to move around and get to where they want to be — the kids can get to sport and school without mum or dad having to drive, which frees up even more time. Reframe the idea that control is out of your hands in an AV, to a message that says you have more control with AVs. Implementing a more emotional ‘angle’ to the control of positive cognition in respect to AVs could potentially increase AV acceptance in Australia. Additionally, more emotionally driven, visual communication such as accidents being avoided in AVs, could also increase acceptance of AVs and help to reduce concerns over safety.
Recommendation 9
Focus/appeal more to individuals’ personal values: a focus on the disabled and aged seems more successful at promoting more acceptance of AVs, as much of the resistance to AVs may come from older parts of the populace — a large, wealthy part; and messages pertaining to a person’s local/region area appeals more to a person’s values. There may be value in reframing around the issue of discrimination against disability/age.
Recommendation 10
That governments and industry should begin the process of preparing for the AV transition of the Australian workforce, and carefully reframe communications. A focus should be on job creation and possibly tackling the unemployment issue, working with tertiary institutions. The shift to automation will be relatively gradual, allowing opportunity for labour to be absorbed by other growth industries. This will be a continual process. In terms of the aging workforce, management of the transition to AVs will need to allow for natural retirement, or retraining into newly emerging roles. There may be scope for reframing around the retention of jobs in Australia, and around Australia not wanting to be left behind in modern technology.
Section 5: Reflection
Group Coordination and Cooperation
Over the course of researching and producing this report, we gained insight into the complexities of the issue of the driverless car. The authors are from somewhat different scientific backgrounds, and so broad a range of perspectives. We each remain, however, in support of the move to driverless cars.
The scoping phase provided a solid entry to our consolidation of the report. Feedback was addressed and incorporated.
One author expressed that there had been some difficulties around deciding on which aspects of AVs we would focus on, given the broad nature of the topic and its wide relevance. There were challenges in our discussion of which risks are pertinent and how they are related — however we seemed to be cohesive and communicating effectively.
It is clear with hindsight that the project group has enjoyed effective and consistent levels of communication between all members, including timely decisions on the division and adoption of project work components. Fast responses to communications and regular group video chats kept everyone on the same page throughout the project.
Project reflection
Media analysis
There are media sources and auxiliary information that may have made our assessment of the campaign more accurate. These include:
- Facebook evaluation on media articles — most of the news articles were also shared across social media when published. On Facebook, there is extra information that can aid in the evaluation of audience responses to each article, such as likes, emoticon responses (sad, happy, angry, amazed) and shares. This can be used for evaluating the accuracy of message responses and message alignment.
- Wider sample size — Our sample size in this instance was quite small, restricted to 5 organisations, and representative of one kind of media source. The communication of driverless cars is spread not just through news articles but other platforms like radio, and social media like Twitter and Youtube.
References
8146.0 — Household Use of Information Technology, Australia, 2016–17. Retrieved March 2018, from: http://www.abs.gov.au/ausstats/abs@.nsf/mf/8146.0
ABC News (18 October 2015) “Driverless trucks move all iron ore at Rio Tinto’s Pilbara mines, in world first” Retrieved April 2018, from: http://www.abc.net.au/news/2015-10-18/rio-tinto-opens-worlds-first-automated-mine/6863814
Abraham, H., Seppelt, B. D., Mehler, B., Reimer, B., Fitzgerald, C., & Coughlin, J. F. (2018). Consumer Interest in Automation: Change over One Year. Agelab White Paper. Massachusetts Institute of Technology. Retrieved March 2018, from: https://www.researchgate.net/publication/322400007_Consumer_Interest_in_Automation_Change_over_One_Year
Anglo American. (2018). Step into our world. Retrieved May 2018, from: http://www.angloamerican.com/
Australian Bureau of Statistics. (2016). Disability, Ageing and Carers, Australia: Summary of Findings, 2015 (4430.0). Retrieved March 2018, from Australian Government website: http://www.abs.gov.au/ausstats/abs@.nsf/Lookup/4430.0main+features202015
Autonomous Solutions, Inc. (2018). Mining | ASI. Retrieved May 2018, from: https://www.asirobots.com/mining/
Autonomous Uber vehicle kills woman pushing bike across street. (2018, March 20). Retrieved April 2018, from: http://www.abc.net.au/news/2018-03-20/uber-suspends-self-driving-car-tests-after-fatal-crash/9565586
AZO Mining. (2018, January 25). Autonomous Vehicles in Mining. Retrieved May 2018, from: https://www.azomining.com/Article.aspx?ArticleID=1381
Bathurst 1000. (2018). En.wikipedia.org. Retrieved 29 March 2018, from: https://en.wikipedia.org/wiki/Bathurst_1000
Bellet, T., Paris, J., & Marin-Lamellet, C. (2018). Difficulties experienced by older drivers during their regular driving and their expectations towards Advanced Driving Aid Systems and vehicle automation. Transportation Research Part F: Traffic Psychology and Behaviour, 52, 138–163. Retrieved March 2018, from: http://dx.doi.org/10.1016/j.trf.2017.11.014
Bogle, A. (2018, March 21). 5 big ethical questions about driverless cars we still need answered. Retrieved April 2018, from: http://www.abc.net.au/news/science/2018-03-21/self-driving-autonomous-cars-five-ethical-questions/9567986
Bonnefon, J. F., Shariff, A., & Rahwan, I. (2016). The social dilemma of autonomous vehicles. Science, 352(6293), 1573–1576.
Bureau of Infrastructure, Transport and Regional Economics (BITRE). (2013). Road Safety of Older Australians — Recent Statistics, Information Sheet 50 (pp. 2–3). Canberra ACT 2601, Australia: Bureau of Infrastructure, Transport and Regional Economics (BITRE).
Burke, L. (2017, August 5). Technology that reduces our road tolls to zero. Retrieved March 2018, from: http://www.news.com.au/technology/innovation/motoring/technology-behind-the-collisionfree-cars-of-the-future/news-story/147dab941040fdf825beab7f41c0331a
Caldwell, F. (2017, September 27). Queensland is preparing for driverless cars to hit the streets. Retrieved March 2018, from: https://www.smh.com.au/politics/queensland/queensland-is-preparing-for-driverless-cars-to-hit-the-streets-20170927-p4yw4w.html
Cars of the future (2018). Retrieved April 2018, from: https://www.9now.com.au/a-current-affair/2018/extras/latest/dec-2017/171228/cars-of-the-future
Chalmers University of Technology. (2018, February 19). Self-driving vehicles for a safer mining industry | Chalmers. Retrieved March 2018, from: https://www.chalmers.se/en/areas-of-advance/Transport/news/Pages/Self-driving-vehicles-for-a-safer-mining-industry-.aspx
Clark, E. (2018, March 20). Uber driverless car crash fails to deter early adopters because ‘humans are inferior’. Retrieved April 2018, from: http://www.abc.net.au/news/2018-03-20/uber-driverless-car-crash-fails-to-deter-early-adopters/9566084
Cowan, J. (2017, March 11). Everything you need to know about driverless cars. Retrieved April 2018, from: http://www.abc.net.au/news/2017-03-11/everything-you-need-to-know-about-driverless-cars/8336322
Defence Advanced Research Projects Agency (2007) DARPA Grand Challenge. Retrieved April 2018, from: http://archive.darpa.mil/grandchallenge/
Demuro, D. (2018). Will Autonomous Cars Kill Car Culture? — Autotrader. Autotrader. Retrieved 28 March 2018, from: https://www.autotrader.com/car-news/will-autonomous-cars-kill-car-culture-260156
Department of Motor Vehicles California (2012), Testing of Autonomous Vehicles, Second Modified Express Terms Title 13, Division 1, Chapter 1 Article 3.7. Retrieved 28 March 2018, from: https://www.dmv.ca.gov/portal/wcm/connect/aa08dc20-5980-4021-a2b2-c8dec326216b/AV_Second15Day_Notice_Express_Terms.pdf?MOD=AJPERES
Donoughue, P. (2015, November 4). To kill or not to kill: Researchers probe moral code for driverless cars. Retrieved March 2018, from: http://www.abc.net.au/news/2015-11-04/researchers-probe-moral-code-for-self-driving-cars/6910606
Faulhaber, A. K., Dittmer, A., Blind, F., Wächter, M. A., Timm, S., Sütfeld, L. R., … & König, P. (2018). Human Decisions in Moral Dilemmas are Largely Described by Utilitarianism: Virtual Car Driving Study Provides Guidelines for Autonomous Driving Vehicles. Science and Engineering Ethics, 1–20.
Foden, B. (2018, April 23). Drivers needed as world-first semi-autonomous vehicle trial launched. Retrieved May 2018, from: https://www.smh.com.au/national/act/drivers-needed-as-world-first-semi-autonomous-vehicle-trial-launched-20180423-p4zb6v.html
Ford Australia. (2018). En.wikipedia.org. Retrieved 28 March 2018, from: https://en.wikipedia.org/wiki/Ford_Australia
Ford: Self-driving cars come with 4 types of hurdles. (2016). Retrieved May 2018, from: https://au.news.yahoo.com/ford-self-driving-cars-come-with-4-types-of-hurdles-32616137.html
Frey, C. B., & Osborne, M. A. (2017). The future of employment: how susceptible are jobs to computerisation?. Technological Forecasting and Social Change, 114, 254–280.
Gerdes, J. C., & Thornton, S. M. (2016). Implementable ethics for autonomous vehicles. In Autonomous Driving (pp. 87–102) Eds. M Maurer, J. C Gerdes, B Lenz & H Winner. Springer, Berlin, Heidelberg.
GM will make an autonomous car without steering wheel or pedals by 2019. (2018). The Verge. Retrieved 28 March 2018, from https://www.theverge.com/2018/1/12/16880978/gm-autonomous-car-2019-detroit-auto-show-2018
Gogoll, J., & Müller, J. F. (2017). Autonomous cars: in favor of a mandatory ethics setting. Science and engineering ethics, 23(3), 681–700.
Gold, N., Colman A.M., & Pulford, B.D. (2014). Cultural differences in response to real-life and hypothetical trolley problems. Judgment and Decision Making, 9(1), 65–76.
Google Official Blog, (2010), “What we’re driving at”. Retrieved April 2018, from: http://googleblog.blogspot.com.au/2010/10/what-were-driving-at.html
Greaves, S. (2018, March 20). The №1 reason we should put driverless cars on the road. Retrieved March 2018, from: https://www.smh.com.au/national/the-no-1-reason-we-should-put-driverless-cars-on-the-road-20180320-p4z57s.html
Haratsis, B. (2016). Economics Impacts of Automated Vehicles on Jobs and Investment (Rep.). Australian Driverless Vehicle Initiative.
Harmsen, N. (2015, November 11). SA hopes driverless car research can reboot its auto industry. Retrieved March 2018, from: http://www.abc.net.au/7.30/sa-hopes-driverless-car-research-can-reboot-its/6920760
Hobbs, B. (2016, June 24). Should your driverless car sacrifice your life? Retrieved Match 2018, from: http://www.abc.net.au/news/science/2016-06-24/driverless-car-attitudes-what-do-we-want/7540512
Holden. (2018). En.wikipedia.org. Retrieved 28 March 2018, from: https://en.wikipedia.org/wiki/Holden
House of Representatives Standing Committee on Industry, Innovation, Science and Resources. (2002). Social issues relating to land-based automated vehicles in Australia. Parliament of the Commonwealth of Australia, Canberra.
http://gs.statcounter.com/search-engine-market-share/all/australia/#monthly-201705-201804
Huebner, B., & Hauser, M. D. (2011). Moral judgments about altruistic self-sacrifice: When philosophical and folk intuitions clash. Philosophical Psychology, 24(1), 73–94.
Hunt, T. (2018). I’m a car enthusiast — bring on self-driving cars. Troy Hunt. Retrieved 28 March 2018, from: https://www.troyhunt.com/im-a-car-enthusiast-bring-on-self-driving-cars/
Hurley, D. (2016, November 8). Self-driving cars a terrorist risk. Retrieved March 2018, from: http://www.news.com.au/national/victoria/why-driverless-cars-could-become-a-terror-threat-to-melbourne/news-story/a94974cf6df652d293263d6270bd47a3
Hutson, M. (2017, December 14). People don’t trust driverless cars. Researchers are trying to change that. Retrieved March, from: http://www.sciencemag.org/news/2017/12/people-don-t-trust-driverless-cars-researchers-are-trying-change
Insurance institute for highway safety — Highway loss data institute. (2016). Older drivers. Retrieved March 2018, from: http://www.iihs.org/iihs/topics/t/older-drivers/fatalityfacts/older-people/2011
ISPOS. (2018). Public opinion on a future with driverless cars. Retrieved April 2018, from: https://www.ipsos.com/sites/default/files/ct/news/documents/2018-04/ipsos_ga_driverless_cars_in_australia.pdf
John McCarthy, (1970), “Computer Controlled Cars”. Retrieved April 2018, from: http://www-formal.stanford.edu/jmc/progress/cars/cars.html
Jurewicz, J. (2015, September 1). Is Western Australia ready for driverless cars? Retrieved March 2018, from: http://www.news.com.au/national/western-australia/is-western-australia-ready-for-driverless-cars/news-story/fa26c7716c0be8e2f0dbdf411c36f0cd
Kahan, D.M. (2010). Fixing the communications failure. NATURE 463, 296–7
Kharpal, A. (2017, June 13). Apple reveals driverless car system to detect pedestrians and cyclists in rare research paper. Retrieved February 2018, from: https://www.cnbc.com/2017/11/22/apple-driverless-car-system-to-detect-pedestrians-and-cyclists.html
Lamberts, R. (2012). Ethics and Accountability in Science and Technology. In John K. Gilbert, and Susan M. Stocklmayer (Eds.) Communication and Engagement with Science and Technology: Issues and Dilemmas — A Reader in Science Communication (pp. 130–147). Taylor & Francis Group.
Leben, D. (2017). A Rawlsian algorithm for autonomous vehicles. Ethics and Information Technology, 19(2), 107–115.
Lewis M. Clements, Kara M. Kockelman (2017) ECONOMIC EFFECTS OF AUTOMATED VEHICLES, The University of Texas at Austin
Liu, H. Y. (2017). Irresponsibilities, inequalities and injustice for autonomous vehicles. Ethics and Information Technology, 19(3), 193–207.
Lloyd, M. (2018, March 23). Driverless cars are now tested on our roads, so should we be worried? Retrieved April 2018, from: http://www.abc.net.au/news/2018-03-23/should-we-worry-about-driverless-cars-being-tested-on-our-roads/9569056
Lowrey, T. (2017, September 11). Job losses, hacking worries and ethical questions ? but get used to driverless cars. Retrieved from http://www.abc.net.au/news/2017-09-11/driverless-cars-job-losses-hacking-issues-and-ethical-questions/8893384
Lungren, R.E., McMakin, A.H. (1998). Risk Communication: A Handbook for Communicating Environmental, Safety and Health Risks. Battelle Press.
Lutin, Jerome M, PHD, PE, AICP; Kornhauser, Alain L, PHD; Lerner-Lam, Eva (2013),”The Revolutionary Development of Self-Driving Vehicles and Implications for the Transportation Engineering Profession,”, MASCE. Institute of Transportation Engineers. ITE Journal; Washington Vol. 83, Issue. 7. Retrieved April 2018, from: https://www.researchgate.net/publication/292622907_The_Revolutionary_Development_of_Self-Driving_Vehicles_and_Implications_for_the_Transportation_Engineering_Profession
Mangalindan, J. P. (2018, January 11). Samsung’s plan to dominate the self-driving car industry. Retrieved from https://au.finance.yahoo.com/news/samsungs-plan-dominate-self-driving-car-industry-170231451.html
Miah, A. (2005). Genetics, cyberspace and bioethics: why not a public engagement with ethics? Public Understanding of Science, 14, 409–421.
Mokyr, J., Vickers, C., & Ziebarth, N. L. (2015). The history of technological anxiety and the future of economic growth: Is this time different? Journal of Economic Perspectives, 29(3), 31–50.
Moore, T. (2018, February 21). Ipswich set to trial electric driverless shuttle buses. Retrieved April 2018, from: https://www.smh.com.au/business/small-business/ipswich-set-to-trial-electric-driverless-shuttle-buses-20180221-p4z145.html
Morgan, M.G. (2002). Risk Communication: A Mental Models Approach. Cambridge University Press.
Murphy, J. (2016, July 1). Is this the end of Tesla driverless cars? Retrieved April 2018, from: http://www.news.com.au/technology/innovation/motoring/the-first-tesla-driverless-death-has-happened-another-one-could-stop-driverless-cars-in-their-tracks/news-story/0ccb24206dc934fbad29ffb8809f0739
National Transportation Safety Board Office of Public Affairs (2017) NTSB Opens Docket on Tesla Crash. Retrieved 12 March 2018, from: https://www.ntsb.gov/news/press-releases/Pages/PR20170619.aspx
Navigant Research (2017), Navigant Research Leaderboard Report: Automated Driving Assessment of Strategy and Execution for 18 Companies Developing Automated Driving Systems Published 2Q 2017. Retrieved 19 April 2018, from: http://62.nl.dealer-preview.co/SiteAssets/LB-AV-17-Navigant-Research_FINAL.pdf
Navigant Research (2018), Navigant Research Leaderboard: Automated Driving Vehicles, Assessment of Strategy and Execution for 19 Companies Developing Automated Driving Systems, Q1 2018. Retrieved April 2018, from: http://62.nl.dealer-preview.co/SiteAssets/LB-AV-17-Navigant-Research_FINAL.pdf
Navigant Research. (2018, January 29). Navigant Research Leaderboard: Automated Driving Vehicles. Retrieved April 2018, from: https://www.navigantresearch.com/research/navigant-research-leaderboard-automated-driving-vehicles
New South Wales Parliament. Joint Standing Committee on Road Safety. (2016). Driverless Vehicles and Road Safety in NSW, Report no. 2/56
News and Media Research Centre. (2017). Digital News Report Australia 2017. Retrieved March 2018, from University of Canberra website: http://apo.org.au/system/files/95161/apo-nid95161-346236.pdf
NRMA. (2017). Trialling Autonomous Vehicles in NSW. Retrieved May 2018, from NRMA website: https://www.mynrma.com.au/-/media/documents/reports-and-subs/trialling-autonomous-vehicles-in-nsw.pdf?la=en
Older drivers — TAC — Transport Accident Commission. (2018). Tac.vic.gov.au. Retrieved 29 March 2018, from: http://www.tac.vic.gov.au/road-safety/safe-driving/older-drivers
One million driverless vehicles by 2035: NRMA — 9Finance. (2017, August 9). Retrieved May 2018, from: https://finance.nine.com.au/2017/08/09/07/36/million-driverless-vehicles-by-2035-nrma
Opray, M. (2018). The end of cars is coming, so what will happen to the petrolheads? the Guardian. Retrieved 28 March 2018, from: https://www.theguardian.com/sustainable-business/2016/mar/22/the-end-of-cars-is-coming-so-what-will-happen-to-the-petrolheads
Parliament of the Commonwealth of Australia. House of Representatives Standing Committee on Industry, Innovation, Science and Resources. (2017). Social issues relating to land-based automated vehicles in Australia.
Pigeon, N. (2014). Risk, Uncertainty and Social Controversy: From Risk Perception and Communication to Public Engagement. In G. Bammer, et al. (Eds.), Uncertainty and Risk: Multidisciplinary Perspectives (349–361). Taylor & Francis Group.
Piotrowski, A. (2017, December 28). What it’s like test-driving Volvo’s driverless car. Retrieved April 2018, from: https://www.9news.com.au/national/2017/12/28/17/01/driverless-cars-in-australia-what-s-the-status
Powers, T. M. (2006). Prospects for a Kantian machine. IEEE Intelligent Systems, 21(4), 46–51.
Quartz, D. K. (2017, August 25). Driverless cars will make a lot of jobs better, not destroy them. Retrieved May 2018, from: https://finance.yahoo.com/news/driverless-cars-lot-jobs-better-103019434.html
Questions asked over driverless cars. (2017, November 24). Retrieved April 2018, from: https://au.news.yahoo.com/questions-asked-over-driverless-cars-37998646.html
QUT | Staff Profiles | Andry Rakotonirainy. (2018). Staff.qut.edu.au. Retrieved 28 March 2018, from: http://staff.qut.edu.au/staff/rakotoni/
Reimer, B. (2014). Driver Assistance Systems and the Transition to Automated Vehicles: A Path to Increase Older Adult Safety and Mobility? Public Policy & Aging Report, 24(1), 27–31. doi:10.1093/ppar/prt006
Reynolds, E. (2016, May 2). The jobs killer is coming. Retrieved March 2018, from: http://www.news.com.au/finance/business/travel/the-jobs-killer-is-coming-how-driverless-trucks-could-change-australia/news-story/4f5b8a42b0452703d62e00f3e7644d7b
Romm, J.J. (2012). Language Intelligence: Lessons on Persuasion from Jesus, Shakespeare, Lincoln, and Lady Gaga. CreateSpace Independent Publishing Platform.
Rothenbucher, D., Li, J., Sirkin, D., Mok, B., & Ju, W. (2016). Ghost driver: A field study investigating the interaction between pedestrians and driverless vehicles. 2016 25th IEEE International Symposium on Robot and Human Interactive Communication (RO-MAN). doi:10.1109/roman.2016.7745210
Sallis, J. (2018, January 4). Driverless cars could be better or worse for our health? it’s up to you. Retrieved March 2018, from: http://www.abc.net.au/news/2018-01-03/driverless-cars-could-be-better-or-worse-for-our-health/9299104
Sandman, P.M. (2003). Stakeholders. Retrieved 3 April from: http://www.psandman.com/col/stakeh.htm
Self-driving bus crashes in first hour of service. (2017, November 10). Retrieved March 2018, from: https://au.news.yahoo.com/self-driving-las-vegas-bus-crashes-in-first-hour-of-service-37816317.html
Shariff, A., Bonnefon, J., & Rahwan, I. (2017). Psychological roadblocks to the adoption of self-driving vehicles. Nature Human Behaviour, 1(10), 694–696. doi:10.1038/s41562–017–0202–6
Smith, A., & Anderson, J. (2014). AI, Robotics, and the Future of Jobs. Pew Research Centre.
Society of Automotive Engineers (SAE), International (2014–10–01), J3016: Taxonomy and Definitions for Terms Related to On-Road Motor Vehicle Automated Driving Systems. Retrieved March 2018, from: https://web.archive.org/web/20170903105244/https://www.sae.org/misc/pdfs/automated_driving.pdf
StatCounter Global Stats — Browser, OS, Search Engine including Mobile Usage Share. (2018). Retrieved April 2018, from: http://gs.statcounter.com/
State of Nevada Legislation (2011) Committee on Transportation, Section 8 and 2, Assembly Bill №511. Retrieved April 2018, from: https://www.leg.state.nv.us/Session/76th2011/Bills/AB/AB511_EN.pdf
Tesla Website (1 April 2018), About Section. Retrieved April 2018, from: https://www.tesla.com/en_AU/about
Tesla Website (1 April 2018), Tesla Autopilot. Retrieved April 2018, from: https://www.tesla.com/en_AU/autopilot
Thaler, R.H., Sunstein, C.R. (2008). NUDGE: Improving Decisions about Health, Wealth, and Happiness. Yale University Press New Haven & London
Timeline: Holden’s history in Australia. (2018). ABC News. Retrieved 28 March 2018, from: http://www.abc.net.au/news/2013-12-11/timeline-holden-history/5150240
Tucker, H. (2015, October 27). I drove a driverless car? it was terrifying. Retrieved March 2018, from: http://www.news.com.au/technology/innovation/motoring/australian-tesla-model-s-owners-can-now-drive-semiautonomously/news-story/f513eb6a7f0e671487b02a3e3e54e569
Video released of fatal Uber self-driving vehicle crash with pedestrian. (2018, March 22). Retrieved April 2018, from: https://www.smh.com.au/world/north-america/video-released-of-uber-self-drive-fatal-crash-with-pedestrian-20180322-p4z5mr.html
Wade, M. (2018, April 5). Australia more cautious about driverless cars than many nations: poll. Retrieved March 2018, from: https://www.smh.com.au/national/australia-more-cautious-about-driverless-cars-than-many-nations-poll-20180405-p4z80g.html
Waymo (Google). (2018). New way forward in mobility. Retrieved April 2018, from: https://waymo.com/
Wehner, M. (2017, June 28). Driverless cars are getting really confused by kangaroos. Retrieved March 2018, from: https://finance.yahoo.com/news/driverless-cars-getting-really-confused-kangaroos-030831974.html
Whigham, N. (2016, June 25). ‘A serious consideration of algorithmic morality has never been more urgent? Retrieved March 2018, from: http://www.news.com.au/technology/innovation/design/the-ethical-dilemma-that-could-pose-a-roadblock-to-the-adoption-of-autonomous-vehicles/news-story/3bd964a801571d1910bc71d1d895c7ce
Williamson, R. C., Raghnaill, M. N., Douglas, K., & Sanchez, D. (2015). Technology and Australia’s Future: New technologies and their role in Australia’s security, cultural, democratic, social and economic systems. Melbourne: Australian Council of Learned Academies.
Zhou, Y., & Tuzel, O. (2017). VoxelNet: End-to-End Learning for point cloud Based 3D object detection. Retrieved May 2018, from: https://arxiv.org/abs/1711.06396
Appendix
The SAE classification system contains six levels of automation separated into two parts based on whether the driver or the system does the majority of the driving, these levels are:
Table 1.0 — SAE Driverless Vehicle Automation Classification System
Level 0
No Automation
The full-time performance by the human driver of all aspects of the dynamic driving task, even when enhanced by warning or intervention systems.
Level 1
Driver Assistance
The driving mode-specific execution by a driver assistance system of either steering or acceleration/deceleration using information about the driving environment and with the expectation that the human driver perform all remaining aspects of the dynamic driving task.
Level 2
Partial Automation
The driving mode-specific execution by one or more driver assistance systems of both steering and acceleration/deceleration using information about the driving environment and with the expectation that the human driver perform all remaining aspects of the dynamic driving task.
Level 3
Conditional Automation
The driving mode-specific performance by an automated driving system of all aspects of the dynamic driving task with the expectation that the human driver will respond appropriately to a request to intervene.
Level 4
High Automation
The driving mode-specific performance by an automated driving system of all aspects of the dynamic driving task, even if a human driver does not respond appropriately to a request to intervene.
Level 5
Full Automation
The automation system performs all driving tasks in all conditions without the need of human drivers.
Figure 1.0 SAE Table of Automation
Figure 2: Google Trends Top Queries (Google Trends).
Figure 4 (Google Trends).
Top Search Terms, Stakeholder Appearance
AV Topics top topics
AV Topic Fastest Topic increase