Introduction
Vehicles were invented over hundred years ago to transport people and goods from one place to another. In the past hundred years the purpose hasn’t changed however increased safety, comfort and efficiency makes them almost unrecognisable today. In the last century autonomous vehicles only existed in people’s imaginations, creating timeless characters on screen from Herbie [1] in 1968 to Johnny Cab [2] in 1990.
In the United States human choices are responsible for 94 per cent of all accidents [3], what if this could be lowered? The discussion of autonomous vehicles is not new however it is becoming increasingly important around the world, for instance the UK’s governments ambition to have “self-driving cars … on UK roads by 2021” [4]. Before this occurs there, are many challenges facing the auto industry and society: What does it mean for businesses? Who is responsible in a crash? Will they really be safer?
This essay will focus on the challenges that society and technology will face for autonomous vehicles in land use on public roads along with all the opportunities. This essay will use the SAE levels of automation [5] as followed:
- Level 0: no automation; the driver performs all tasks
- Level 1: driver assistance; the vehicle is controlled by the driver, but driver assist features may be present
- Level 2: partial automation; vehicle has combined automated functions, but the driver must remain engaged and monitor the environment at all times
- Level 3: conditional automation; driver must be ready to take control of the vehicle at all times but is not required to monitor the environment
- Level 4: high automation; the vehicle is capable of performing all driving functions under certain conditions. The driver has the option to control the vehicle
- Level 5: fully autonomous; the vehicle is capable of performing all functions and the driver may have the option to control the vehicle, but no driver is required
Furthermore, this essay will define vehicle as “a machine used for transporting people or goods on land” [6]
Technological challenges
In order for a car to be autonomous, the car will need sensors to collect information about the road and a central processing unit to analyse all the data and make decisions appropriately. There will need to be a range of sensors (see figure 1) these include [7] lidars, radars, cameras and ultrasound to recognise the surroundings and measure the distance between the vehicle and nearby objects so the central processing unit can evaluate them and respond accordingly. The vehicle will also need sensors to assess the condition of the car, for instance wheel revolutions and engine. Although this technology already exists there are many problems with identifying objects. For example, rain, snow, road works, complex city driving, and other obstacles can prevent the vehicle from carrying out the right driving action.
The next big challenge for technology is the software and algorithms that can learn as vehicles move more so the system is able to evaluate unknown data and respond to events it has not been programmed for as well as make real time decisions about unforeseen situations. In order to work efficiently the vehicles will have to be able to communicate with each other (V to V communication) via an ad-hoc network (a temporary connection between vehicles in close proximity). This means that vehicles will be constantly transmitting information on direction and speed of their journeys within a range of 300-1000m, for example a car driving into a traffic jam would be constantly deaccelerating and sending signals. All vehicles behind would detect these signals and decrease their speed or alert the driver to do so. It is believed that the technology will reduce crashes in situations that do not involve human drivers In Ann Arbor, Michigan, USA, a test has been running since 2012 with 3,000 vehicles equipped with V-to-V communication. As well as providing data this test is hoping to improve safety of road users in the coming future
Vehicles will also need to communicate with infrastructure (V to I). This will mean upgrading every traffic sign, so it is equipped with a transmitter constantly sending information on its location and message and upgrading road markings, so they are visible to both humans and machines in all conditions. This will require a significant investment from governments. The parts of the UK are being upgraded to “smart motorways” [8] which use technology to actively manage the flow of traffic. This will be the first step in creating the environment for autonomous vehicles. Both V-to-V and V-to-I communication will need full 5G coverage, Highways England is exploring improving [9] Wi-Fi and 5G connectivity, as part of the governments ambitions to have “self-driving cars… on UK roads by 2021” [3]. For real time V-to-V communication, LTE-V (Long term evolution- vehicles) communication technology is being compared to WLAN 802.11p in various situations by Audi on the A9 autobahn in Germany, allowing the industry to specify the requirements for standardisation of future 5G communication [10]. This will mean that autonomous vehicles will collect a massive amount of data which could be used to create personalised services and products. This makes it essential that the right balance is found between sharing public and private data.
This software brings with it the risk of hacking, perhaps making them the perfect tool for terrorists. In theory hackers could hack vehicles, kidnapping, or deliberately causing an accident could occur. Furthermore, this is a new problem for the auto industry which they have got a lot of experience in compared to companies like Apple or Google. these problems could change people’s perception, thus making people unwilling to embrace this new technology. To prevent these problems the auto industry will have to invest heavily in cybersecurity.
Impacts on Economics and business
According to a study done by Morgan Stanley Research in 2013 [11] in the united states $1.3 trillion could be saved with autonomous vehicles, these savings are mainly from reduced fuel consumption, improved productivity of the vehicle’s occupants, reduction in accidents and the related cost of accidents.
A clear benefit of autonomous driving is the time people can gain. In London, 2016 the average commute was 46 minutes [12]. Long commutes are shown to decrease productivity and has a negative impact on mental health, [13] as it was reported workers with a long commute are 33 per cent more likely to suffer from depression. With an autonomous vehicle, people would no longer need to drive so this time could be used exercising, responding to email or watching a film to name a few. Providing these services in the car would give new business opportunities. Also, this increased mobility will allow people to live further away from their place of work and permit people to find better jobs due to increased mobility. Furthermore, driver assist programs that allow vehicles to park in garages without a driver in the car are already reducing the space requirements, creating space which could be used for playgrounds, parks and sports grounds.
The traditional business model of a car manufacturer is changing as there will be less emphasis on hardware and more emphasis on software. Furthermore, traditional car manufactures are not the only players, companies like Google, Apple and others are also entering the field. In order to survive car manufactures must be willing to adapt, to develop a new business model
An additional aspect will be more car-related services, parking and increased mobility and connecting the car with the environment including the internet. Also, entertainment in the vehicle such as cinema, audio or apps. Another source is the volume of big data (vehicle, passenger and location data) which could be used to offer travellers access to data-based services. These provide huge economic opportunities as well as an entrepreneurial challenge for the auto industry as the structure of car manufactures business will change and they will have to give up expertise which has been gained for over hundred years.
Like with any technology, autonomous driving will lead to job losses, the most evident being truck drivers and taxi drivers. Using data from the 2010 to 2014 merged American community survey it is predicted more than four million jobs will likely be lost [14]
But many jobs will be created in different sectors, such as software development and engineering as well as cybersecurity and data management sector; autonomous vehicles will produce a massive amount of data which will need to be manged and protected. Also, entertainment and marketing sector; the time wasted driving passengers will be able to utilise watching tv, sending emails et cetera.
autonomous vehicles will require a new insurance policy although autonomous vehicles will be safer therefore accidents will occur less frequently. This will put more pressure on the insurance company’s business model. However, when accidents do occur, they will be at a greater scale therefore the effects will cost more. In the United Kingdom the Department of Transport have announced plans to establish a single insurer model where an insurer covers both the drivers use of the vehicle and the autonomous vehicle technology. [15]
Law and government challenges
The arrival of autonomous vehicles poses many legal questions, in particular who is responsible in an accident? In level 5 the driver has no control therefore many are suggesting that the manufactures, designers and software developers may face liability [16] however if the autonomous vehicle is to be self-learning it is impossible to know whether the robot’s actions are because of what is independently learned or original programming. This raises the question can you blame the manufacturer if the vehicle learnt to do it?
As of 2019 many countries have taken steps to create and impose legalisation for testing autonomous vehicles on public roads. In 2017 Germany passed laws allowing companies to test self-driving cars on public roads on condition a licensed driver is behind the wheel to assume control if necessary [17]. China and California have followed, as China has released national guidelines on road tests on condition the vehicle passes a technical assessment and is equipped with sensors and cameras so the driving position can be monitored [18]. These new laws allow companies such as Cruise and Apple conduct tests which help supply information for further development of V-to-V and V-to-I communication.
Autonomous driving will require national law to be fabricated as well as international law, such as the 1968 Vienna convention [19] will need to be modified as stimulates that a human driver must always remain in control of the vehicle and is responsible for their behaviour in traffic.
Ethics and public perception
One of the biggest challenges for autonomous vehicles is public perception, if autonomous vehicles are labelled disruptive like genetic engineering, which took years before proper arguments were possible. Or like nuclear power, where the discussions about what to do about nuclear waste are occurring after the technology was established. Both of these show that resistance in public opinion can bring technology to a standstill. To prevent this from happening the industry and government will need to discuss the issues surrounding autonomous vehicles openly and honestly. The current discussion in the media shows that people are aware of the possible dangers but there is less emphasis on the potential benefits, for instance in 2016 [20] a tesla on autopilot drove a man 20 miles to hospital, saving his life.
A central aspect to autonomous vehicles is ethical principles, one example of this is the trolley experiment, which is based on a theatrical thought experiment [21]. Should a car swerve towards a couple of people or a large group of bystanders? In 2016 MIT conducted a moral machine study and found that 50 per cent of people would want their driverless car to plough into pedestrians rather than harming passengers, furthermore the majority of people would save the lives of humans over animals and spare the lives of many rather than a few. When level 5 automation occurs and is widespread these kinds of problems could occur at any time and the decision must be made by pre-programmed algorithms therefore ethics must be programmed in advance and these moral decisions must be made by people. This could potentially lead to legal issues as it rises the questions in this situation would the person who programmed the ethics be responsible? Some of these questions are already being addressed, Germany has created the first ethical rules which states “In the event of unavoidable accident situations, any distinction based on personal features (age, gender, physical or mental constitution) is strictly prohibited. It is also prohibited to offset victims against one another”. [22] this provides a good start and allows German car makers to move forward with their plans however more will need to be done as international laws must be adopted.
Conclusion
Whatever happens in the future one thing is certain, autonomous vehicles will fundamentally change the way we travel. As people move away from the drivers seat to the passenger seat. What once only appeared on the silver screen as Herbie [1] or Johnny Cab [2] is emerging into reality. The challenges for society and politics, creating legalisation and ethics that will govern the auto industry as it designs and constructs the future of automobile transport.
...(download the rest of the essay above)