Autonomous Vehicles

From New Media Business Blog

Jump to: navigation, search

Contents

Introduction

An Introduction to Autonomous Vehicles

Technology Used

Autonomous vehicles use several types of technologies:

  • Laser Illuminating Detection and Ranging (LiDAR) Sensors: LiDAR sensors maps objects in 3D by bouncing pulses of light off the vehicle’s surroundings.[1] It helps detect the edges of roads, identify lane markings, pedestrians, and other vehicles.[2] LiDAR is the most precise sensor, firing off millions of beams of light per second in a 360-degree way.[1] Although LiDAR is very precise, it is a bulky and expensive equipment.
  • Radio Direction and Ranging (Radar) Sensors: Radar sensors is placed around the vehicle and utilizes radio waves to measure distance of the obstacles.[1] It is less accurate than the LiDAR sensor, but it can work in most environmental conditions.[3]
  • Ultrasonic Sensors: Ultrasonic sensors is an acoustic wave with a very high frequency where humans cannot hear it.[4] Its initial purpose is to detect the immediate environment for parking assistance but is now being used in autonomous vehicles for automatic parking.[4]
  • Cameras: Cameras detect traffic lights, read road signs and keep track of other vehicles, while also looking out for pedestrians.[1] Cameras do not function well in bad weather conditions and in conditions with low light or glare. [5]
  • Central Computer: The central computer in the autonomous vehicle analyzes the data from the various sensors and connectivity to manipulate the steering, acceleration, braking, and route guidance. [2]
  • GPS: It is important to have a highly precise GPS system to ensure that the autonomous vehicle operates safely and is positioned on the road accurately.[6] Vehicles rely on detailed maps of streets to navigate to the location that the driver set it to.

Autonomous Vehicles Level

The following are levels in the autonomous vehicle defined by the Society of Automotive Engineers (SAE) International. SAE International is a global association and is the world’s largest automotive standard setting body:[7]

  • Level 0 - No automation: The vehicle is not automated in any way and relies on a human for all driving tasks such as steering and braking. The vehicle does not provide any assistance to the drivers at all. Most vehicles on the road today fall in the level 0 category, even if the vehicle has a blind-spot monitor feature.[8]
  • Level 1 - Driver Assistance: The driver and computer systems work together to move the vehicle on the road but most of the work is still done by the driver. The vehicle may have one or more system that can control the vehicle’s speed or steering but it cannot operate simultaneously.[8] It can only perform a single task. Some examples of these features include:
    • Adaptive Cruise Control: This is a system for vehicles that automatically adjusts the speed to maintain a safe distance from vehicles ahead.[8]
    • Automatic Emergency Braking: This feature can sense the vehicle ahead and apply the brakes if the driver does not respond. [9]
  • Level 2 - Partial automation: The steering and speed of the vehicle are controlled simultaneously by one or more driver assistance systems. The vehicle can maintain its speed, slow down to avoid other vehicles, or stay in its own driving lane.[10] Although the vehicle can control some functions, the driver still needs to control other elements of driving and monitor the environment surrounding the vehicle. Currently, Tesla vehicles are in level 2 autonomy as they feature the functions outlined above. [11]

Vehicles in level 3 and above are considered automated driving systems:

  • Level 3 - Conditional automation: Vehicles are constantly monitoring the driving environment around them and making decisions themselves. [10] For example, it can see a slower moving vehicle in front before deciding to overtake it. The driver only intervenes if the systems are not functioning properly when the weather confuses the vehicle’s sensors or when the painted lines on the roads are not visible. [10] Drivers should still know when the take over the vehicle if the systems fail. [8]
  • Level 4 - High automation: In level 4, no driver interaction is needed and the vehicle can deal with any situations by itself.[10] If something goes wrong, the vehicle can stop itself when the systems in the vehicle fails. For example, the vehicle will stop on the road and engage hazard lights.[8] However, there may be some geographical limitations on where the vehicle can go such as a city downtown, where it can be difficult to navigate through the busy roads.[10]
  • Level 5 - Full automation: No human control is needed at all because the vehicle is fully automated and can handle all driving tasks in any environmental conditions. [12] There are no pedals, steering wheels, or controls for a driver to take control. The driver just needs to tell where the vehicle should go. An example of this would be Waymo's Firefly as the vehicle has no steering wheel, no pedals, and only required a preset destination inputted to it's software. [13] Similarly in a recent article, Elon Musk has announced his plans to enhance Tesla's vehicle features to upgrade it from level 3 to level 5 autonomy in two years. [11]

Advantages

Safer Roads

Each year, many people are injured or killed from vehicle accidents due to human errors such as fatigue, drinking and driving, or other distractions around them.[14] With the reduction of these factors interfering with driver judgments, injuries in America are expected to fall approximately 4.6 times, from 6 million to 1.3 million. Deaths in America are also projected to decrease 2.9 times, from 33,000 to 11,300. Once the technology is advanced enough, the use of autonomous vehicles is beneficial for situations where there is poor visibility as it can detect small obstacles on the road.

Incompetent drivers or those who are not diligent can also be deterred from the rise of these vehicles. Basic driving essentials such as parking and checking for pedestrians can be done through machinery. Atrocious parking will no longer be an issue and parked cars in the environment will be safer overall since software can determine how to perfectly park a vehicle. Surrounding pedestrians can also have a peace of mind due to the ability of autonomous vehicles to fully obey traffic signals and having an instant reaction to emergency braking systems.

Traffic Efficiency

With the use of autonomous vehicles, there will be less traffic congestion. Traffic efficiency will greatly improve as a result of the reduction or elimination of traffic accidents. Furthermore, speed limits can be increased and recommended safety gaps intended to account for human error can be reduced to adjust for the added safety benefits provided by autonomous vehicles.[15] If autonomous vehicles become advanced enough to be automatically sent to locations to pick-up passengers, families will only need to own one vehicle. The difficulty to coordinate schedules will be alleviated. Vehicle utilization time can be increased and as a result, there will be fewer vehicles on the road. These factors together will allow individuals to not experience being stuck in traffic for long durations and increase their time efficiency.[15]

Fuel Efficiency

Since software will be able to calculate the optimal level of braking and acceleration and complete trips using the shortest route to a destination, fuel consumption will decrease. It is estimated that a widespread adoption of autonomous vehicles will reduce CO2 emissions by approximately 300 million tonnes per year.[15] The positive impact on the environment makes this a very attractive technology. Additionally, the fuel efficiency will also allow consumers to save on fuel costs, making the investment in these vehicles very enticing.

Individual Benefits

Individuals can enjoy more leisure time as they will no longer need to spend time in traffic and commuting to destinations. This opens up free time where people can do other activities such as entertainment, travel overnight while sleeping, and do work. Analysts estimate that worldwide, 1 billion hours of commuting can be saved.[15] It is also beneficial to those who have difficulties driving such as disabled individuals or those who have health issues that hinder them from driving. Additionally, elderly people, young children, and those who are visually impaired will no longer require other people to drive them to desired locations.

Overall, each individual vehicle's safety will also be improved as there will be less car theft and vehicle maintenance can be tracked.[15] Since vehicles will be able to monitor its own surroundings, this self-awareness technology allows for better security detection. Cars can sound the alarm, drive to another location, or completely shut itself off if it detects suspicious activity. Software can also be utilized so that the user can be notified of required vehicle maintenance much quicker made possible by the use of electrical components. This not only reduces check-up costs but also improves the possibility of vehicle breakdowns or failure, thus reducing accidents.

Complications

Companies will have to consider the following aspects when designing and building autonomous vehicles.

Technology

Computer Malfunction

If a computer malfunctions, such as minor glitches, it can cause vehicle accidents. Problems arise when the vehicle is unable to read traffic signs which can also lead to accidents.[15] An example of this is in 2016 when a Tesla car in autopilot mode crashed, resulting in the death of Joshua Brown. The software in the car failed to distinguish between the bright sky and a large white 18-wheel truck and ended up driving the car into the trailer of the truck. [16]With the death of Joshua Brown, computer malfunction has been a big concern with consumers as they start questioning the ability for a computer system to ensure their lives are safe and secure.

Software Security

CMCV.
Charlie Miller and Chris Valasek.

There are concerns that hackers can get into a vehicle’s software and control the vehicle. It can be frightening as to what hackers may do with the vehicle such as crashing the controlled vehicle into buildings or other vehicles. An example of this happening was when Charlie Miller and Chris Valasek tested a car-hacking research they have been working on for the past year. They hacked a Jeep Cherokee in which they controlled the air conditioning, radio setting, windshield wipers, and even cutting the transmissions. [17] Hackers can use software systems that allows them to send commands through the Jeep’s entertainment system to its dashboard functions, steering, brakes, and transmission all through a laptop that can be located anywhere in the world.

In addition, there may be some major privacy concerns. Hackers may be able to steal data from the vehicle to understand where the individual is travelling or even access personal data. With this stolen data, hackers can study an individuals routine and utilize this information with malicious intent.

GPS

There may be some difficulties getting signals in some areas. This would be a concern for the autonomous vehicles that rely on GPS signals to reach their destination. Autonomous vehicles may be confused about their current location and veer off the correct route to the final destination.

Safety

Weather Conditions

Extreme weather conditions include heavy fog, rain, and snow. This interferes with roof-mounted laser sensors and confuses LiDAR sensor and cameras as signals can bounce off the rain or snow.[18] This may result in the difficulty in detecting the lane markers and other markers that help the vehicles to drive safely. If the lanes are not visible or there are unmarked roads in remote areas, the vehicle cannot drive or change lanes safely.

Altering of Road System and Infrastructure

Autonomous vehicles may have to deal with unusual encounters and decide how to react to it.[19] For example, when the stoplights do not work and a traffic officer is directing the traffic. It is relatively simple for drivers to understand construction zone, but it is difficult for autonomous vehicles to interpret it. Road constructions can confuse autonomous vehicles because they are so dynamic. The appearance of a construction zone can vary due to location or types of construction with the different road signage, cones, and workers giving hand signals.[20] An autonomous vehicle can also be confused from a lazy hand wave of a construction worker to a hello wave from a colleague.[20] In addition, it is difficult for engineers to write codes to instruct what the vehicle should do in these situations.

In order to solve this problem, the government and private companies can document active construction zones so that the vehicle navigation systems can avoid these areas .[20] If the vehicle is not fully autonomous, the driver will have to take control and be aware of the surroundings.

Ethical Challenges

Safety Issues

How safe should the autonomous vehicle be? Autonomous vehicles will have to start embedding moral principles by creating different scenarios that the autonomous vehicles must decide on because it does not have human decision making powers.[21] They will have to decide on what safety options they will forgo and the trade-offs made.

An interesting question to ask is: if you were an owner of an autonomous vehicle or if you were just riding in one, would you rather have the vehicle save the pedestrians or save yourself? Should it be programmed to crash into something rather than run over pedestrians, even if that meant killing the vehicle’s passengers?


Group Thoughts: We believe that we would buy a self-driving vehicle that ensures the safety of the driver even if it meant killing the vehicle’s passengers. Drew presented us with a question of whether we would crash into a brick wall or run over two pedestrians in order to save ourselves. It is a difficult question to answer. A team member proposed that they would rather have crash into a brick wall because if they were to run over the pedestrians, they will be faced with consequences. These consequences include that the driver will remember how they injured or even killed the pedestrian for the rest of their lives. This may lead to psychological effects due to the traumatic experience they faced. Family members of the pedestrians may also blame the driver for their mistakes.

Mistakes

People are more inclined to forgive mistakes made by human than machines.[22] People can lose faith in a machine quickly when it makes mistakes therefore, people may not adopt the new technology at all. For example, autonomous vehicles can save the lives of thousands of motorists but can cause fatalities of cyclists and pedestrians to increase which will erode the public’s trust in the technology. If a drunk driver caused an accident, they can easily be sent to prison, but if there are no drivers, then who is there to blame?

Who is responsible if the vehicle crashes?

  • The one who developed it?
  • The owner of the car?
  • The manufacturer of the car?


Group Thoughts: The responsibility depends on the level of autonomous of the vehicle. At the level 0 of autonomy, the driver should be responsible if the vehicle crashes. However, if there is a mechanical issue, then the one who developed is responsible. For level 1 and level 2, we believe that everyone has a joint responsibility if the vehicle crashes. When the level of autonomy is above 3, then the responsibility lies with the one who developed it. The reasoning is that humans interact less with the vehicle's control and relies more on the technology provided by the developer.

MIT's Moral Machine

In the Moral Machine, you, as the moral observer, decide who lives and who dies in 13 given scenarios. The Moral Machine was developed by a group at the Massachusetts Institute of Technology (MIT) and is a way to learn how humans make decisions that would eventually contribute into machine intelligence for advanced technology such as autonomous vehicles.[23]. Based on the decisions made, the machine will measure the demographics you value most (e.g. women vs men, old vs young, the wealthy vs the poor) in comparison to others who went through the same test. At the end, you have the option to reject or take part in the research by allowing MIT to obtain the decisions made during the process. The objective of the platform is to gather information on the decisions made when faced with two catastrophic outcomes. This type of simulation plays a crucial role in the development of self-driving vehicles since, very often, there is no right or wrong answer. The research gathered can be used to guide the self-driving vehicles through challenges and how it should react when faced with similar life-and-death situations.[23]

mit.
MIT's Moral Machine.

Autonomous Vehicles Today

Autonomous vehicles are gaining traction as companies realize the potential and opportunity in the industry. Recently in 2017, many companies have started to declare their intent to create autonomous vehicles and self-driving technology. Additionally, a plethora of companies that are already in the industry have started to announce new innovations and creations for their own fleet of cars. These are exciting times for the autonomous car industry as it is starting to pick up with companies creating new technologies and features to gain a competitive advantage in the market. We will be looking at some of the popular companies in the industry, including Waymo (Google), Tesla, UBER, and General Motors. We will analyze their history and how their autonomous cars work.

Waymo (Google)

When Waymo was still operating under the Google brand, the company started transforming autonomous vehicles from being a vision to being a reality. In 2009 the company modified a Toyota Prius by equipping it with various cameras and sensors and tested the vehicle on the road. The trial was a success as the car managed to drive autonomously for 100 miles.[24] This monumental achievement is what fueled the momentum and interest in the creation of autonomous vehicles at Waymo (Google).

Firefly

A picture of the Google Firefly.
The Google Firefly.

Waymo was one of the very first companies to create a Level 5 autonomous car. Their infamous Firefly car was a fully autonomous car that had no steering wheel or pedals. With no human steering ability, the car relied on GPS technology to get to it’s set destination. Despite the full autonomy, the car could only go up to a maximum speed of 40km/hr. [24] However, the Firefly project has been cancelled as of June 2017 as Waymo wants to focus on their current fleet of cars. Despite its cancellation, a major positive takeaway for our society from the Firefly project is that this could be the start of more companies taking a look at Level 5 autonomous cars.

How it Works

In order for Waymo’s autonomous vehicles to work, they rely on numerous technologies that include sensors, lasers, radars, cameras, and software. Using advance sensor technology and software, their sensors are able to detect human behaviors. For example, when a bicyclist fully extends their left arm out, this is a universal sign that the bicyclist is going to turn left. The car’s sensors will be able to detect this behavior and the software will predict that the bicyclist is going to turn left. Upon recognition of human behavior, the software will adjust the car’s current driving state to make it safe for both parties. In this case, the software will instruct the car to slow down to give the bicyclist more room to change lanes. [25] Regarding their laser technology, its LiDAR hardware is able to generate 2.2 million points per a second. These points are then used to generate a 3D map of the cars environment which will be used for comparison through the vehicle’s software. Lastly, their autonomous vehicles have four radars placed on its bumpers to allow the vehicles to detect items farther away in environments where cars are moving quickly. Additionally, it has a camera designated to detect traffic lights near the rear-view mirror. [24]

The Brain

The main operator behind Waymo’s autonomous vehicles are the very detailed maps created beforehand. Prior to test drives, Google engineers go to the routes to collect data regarding the environment and roads. With the data collected, maps are created and loaded into the car’s software. This means that a virtual world is created showing what the environment and roads are expected to look like if nobody were on the streets. Ultimately, Waymo’s autonomous vehicles are not mapping their environment on the go and from scratch but rather comparing its current environment to the virtual world. Having the ability to compare the differences, the software is able to determine which items are not stationary and can then instruct the car to adjust based on these differences. [26]

Tesla

 A picture of what Tesla cars detect
A comparison of what you see and what Tesla's software sees.

Through software updates, Tesla is in the implementation stage of its autonomous driving features. Compared to other companies, Tesla doesn’t rely on LiDAR technology for its autonomous vehicles to function. In contrast, Tesla relies on 12 ultrasonic sensors, eight surround cameras, radars, and its computer software Tesla Vision. [27] As we can see, the car is heavily based on its cameras and sensors. Using these technological features, Tesla cars are consistently scanning objects as it drives along and it can also use other car positions and road markings to position itself. The image on the right is an example of what Tesla Vision sees based on its technological features. Below is a breakdown of what Tesla Vision detects:

  • The green squares represent objects that are in the path of the car.
  • The blue squares represent objects around the car.
  • The red and pink lines represent road lane lines.
  • The orange squares represent traffic lights.
  • The purple squares represent road signs.

By being able to detect these items, Tesla cars can drive autonomously which leads to some unique features that they offer right now and in the future.

Autonomous Features


Summon

With Tesla’s summon feature, users can summon their car to a spot with the click of a button on their phone app. Currently, the summon feature can have the car turn on, open the garage door, back out to a designated spot on the driveway, close the garage door, and then adjust all the preferences inside the car with the click of a button. [1] This makes mornings easier for people as the car is already prepped and ready to be driven away.

Full Self-Driving Ability

A distant vision Tesla has incorporates Summon with full self-driving ability and hands free autopilot. Tesla has plans for the future where when you enter your vehicle, you can communicate with it and tell the system where you want to go. The car will then proceed to drive you to that location on its own and without any human interaction. Additionally, if a user enters the car and doesn’t tell the system where they want to go, the system will look at the user’s schedule to determine where the car should set its destination. Upon arrival at the destination, the user can exit the car and the system will automatically find a parking spot on its own. When the user is ready to leave again, all they have to do is summon their car and the car will arrive within steps of the user. [1]

Uber

 A picture of the Otto Truck
The Otto truck with autonomous vehicle hardware.

Like most companies in the autonomous vehicle industry, Uber is still in its testing stages. Currently the company is testing in Pittsburgh, Tempe, and San Francisco. Much like Waymo, Uber uses the same technology to power its autonomous vehicles. However, the equipment on Uber’s autonomous cars are noticeably bulkier and bigger. A major contributor to the bulkiness of the equipment on Uber’s autonomous vehicles is due to the 20 cameras it is equipped with to detect objects. By having an abundance of cameras, the car can safely detect objects that it needs to avoid and objects that it should adjust its driving state for. With LiDAR technology that beams out 1.4 million lasers per second, the software in Uber’s autonomous vehicles can generate 3D maps from scratch that shows a cars' current environment. This map is then used to determine where the autonomous vehicle can safely move to. Finally, Uber also has a camera designated to detect traffic lights so it knows when to proceed through an intersection. [2] However, one major difference between Uber and the other companies is that Uber has expanded further into the autonomous vehicle industry and is not just focusing on cars. Uber successfully created and tested a self-driving truck called Otto. Otto was a level four autonomy vehicle that successfully traveled 120 miles on a highway to deliver a freight of beer to Colorado Springs.[3]

General Motors (GM)

 A picture of GM's autonomous vehicle
The Chevy Bolt EV with autonomous vehicle hardware.

As of June 2017, General Motors has become the first company to mass produce autonomous cars.[4] Although GM has the ability to mass produce these cars, the company is still currently in its testing stage. Their autonomous vehicles have been deployed in Scottsdale, San Francisco, and Detroit and plan on releasing 300 more test cars on the roads. [5] GM has been tight lipped with how their autonomous vehicles will work but based on the photos of their test cars, we assume they will follow the current industry standard hardware and software. This includes having radars, laser scanners, high-res cameras, sensors, and software equipped in their autonomous cars.

The structure GM has in building their autonomous cars and autonomous driving software is unique and different compared to other companies in the industry. In 2016, GM bought a company called Cruise Automation. Since the acquisition, Cruise Automation has solely focused on the software portion of autonomous driving. Despite being acquired by GM, Cruise Automation acts as a separate company from GM and is not controlled by GM nor integrated into GM’s operations.[6] By having Cruise Automation as a separate entity from GM, the company is able to focus on advancing autonomous driving software and not be influenced by GM’s own agenda and plans. Currently Cruise Automation has a plan on merging LiDAR, radar, and visual information to create HD maps that are accurate to the millimeter. [7] By having Cruise Automation focus on autonomous driving software, GM has the ability to focus on their own fleet of autonomous cars and hardware. Currently GM has been focusing on their fleet of Chevy Bolt EVs and the hardware that is equipped on it to make autonomous driving possible.

The Rise of a New Industry?

With Apple announcing their entrance to the automated vehicle sector, it raises the question of whether or not the automated vehicle industry will break off into two different sectors. The difference between Apple and other companies in the industry is that Apple announced they are focusing solely on the software that power autonomous vehicles. This creates a debate on whether or not other companies will follow Apple’s footsteps and cause a split between companies focusing on building a complete autonomous vehicle and companies focusing only on the autonomous driving software. We already see this split occurring with GM and their subsidiary company Cruise Automation. GM focuses on building the hardware and cars used for autonomous driving whereas Cruise Automation focuses on building the software to be used by GM's autonomous vehicles. This is exciting for the future because we may see partnerships, that we would have never expected to occur, between vehicle and technology companies. However, with partnerships in the industry, it increases the concerns and complexity regarding accountability when mistakes occur as stated earlier.

Duckietown

Welcome to Duckietown: fake ducks, real science

Duckietown is an advanced autonomy class that originated from MIT during 2016. The class provided students with a low-cost learning environment by allowing them to create autonomous vehicles. These vehicles would then be placed in a miniature town (called "Duckietown") for testing. All content produced by students will be open-source so other students can build on each others work. In addition to educational purposes, this program aims to eventually have a developed, capable software to be placed in worldwide car manufacturers, such as Honda. As of 2017, six universities across the world has adopted this program with various others that uses the open-source platform.

This class opens the opportunity to test various real-life scenarios and factors in a low-cost, controlled environment. One example is that road signs can be tested for visibility and readability. Students have found that ducks' beaks have similar colour tones compared to a stop sign so their software executed halts whenever it detected ducks on the road. Since these vehicles only cost approximately $100, the low-cost nature of unfavorable outcomes allows students to test crash scenarios. Exit strategies and how to minimize impact can be implemented into real-life applications.

Autonomous Boats and Aircrafts

Although there is a lot of emphasis on companies who are designing autonomous cars and trucks, autonomous boats and aircrafts are emerging. Currently BAE systems, a global defence, aerospace and security company, is testing out driverless boats and aircrafts in England [1]. Autonomous boats and aircraft can be beneficial to the defence and oil industry.

Massachusetts Institute of Technology (MIT) and the Amsterdam Institute for Advanced Metropolitan Solutions (AMS Institute) have collaborated together for the the "Roboat" project [2]. The Roboat is currently being tested in canals of Amsterdam for a five year research project [3]. A Roboat has multiple functions including transporting goods, people, and for creating temporary floating infrastructures [4]. It can also monitor the level of pollution by evaluating the water and air quality. If Roboats does become successful, it will be used around the world.

Impact on Industries

Public transportation / Ride-Hailing Services

The introduction of ride-hailing services, such as Lyft and Uber, is thought to reduce the number of users that take public transit. However, research shows that ride-hailing services complement public transit instead. The reason for this is because of the higher ride-hailing fares that are associated with Uber and Lyft. Unless the convenience or efficiency of taking an Uber or Lyft outweigh high costs, users rather choose public transit if it is an option [5]. Ride-hailing services complement the fixed routes and schedules of public transit. It is a method to get to a public transit route and a method to get to the desired destination that is not along the fixed route. By making public transit more accessible via ride-hailing services, we could see a decreased desire in owning a car or owning fewer cars in general [5].

Currently, ride-hailing services are operated by human drivers that drive the rider from one destination to another. The human factor in the ride-hailing service opens doors to safety concerns for the rider and the driver. It also introduces additional problems such as tipping or denying customers due to weather, traffic, or the condition of their surroundings. Introducing autonomous vehicles in the ride-sharing industry will remove the human interaction portion of the service and eliminate human emotions of denying customers due to environmental or safety concerns. The trends of both ride-hailing services and public transit coexisting are still relevant. One major difference is that labor costs will significantly reduce because they are no longer required. As a result, we may see a decrease in ride-hailing service fares. Since each individual car requires a cost for maintenance due to longer operating times and wear and tear in general, we do not see service fares sinking below public transit fares. Also, because of possible congestion on the roads, public transit such as subways or sky-trains with fixed routes may get the rider to their destination more efficiently. Once again, automated ride-hailing services are a complement to public transit, not a competitor.

Insurance

Over 90% of accidents are caused by human error each year, which is why insurance companies like ICBC increase the cost of insurance every year. Insurance rates are currently calculated by the behavior of drivers. Drivers who follow rules and exercise accident avoidance on the road get a discounted rate for insurance. The introduction of autonomous vehicles will change the insurance industry because if drivers do not have control over the car’s performance, the driver is no longer responsible for the actions caused. We will see the insurance model shift from driver and owner liability to automobile manufacturer liability [6]. For example, when buying insurance, the company will check the purpose of the vehicle and ride distance rather than the user’s driving experience and records. Because autonomous vehicles will reduce the number of accidents caused by human error per year, we may see ICBC decreasing the cost of insurance for its customers [7] Currently, the technology is still too new to pass concrete legislation of who is liable for an accident involving an autonomous vehicle. Therefore, as of now, accidents will be reviewed on a case by case basis.

Media and Entertainment

Once the autonomous vehicle is fully integrated into the city, a car can drive itself to and from a designated location with a click of a button. According to the American Driving Survey, the average American spends 46 minutes on the road daily [8]. With 46 extra minutes a day, mainly consisting of driving to and from work, passengers are allocating the time saved towards other activities. For example, the extra time could be used for recreational purposes such as an increased consumption of media and streaming services like Netflix and Hulu [9]. We see passengers garnering a larger social media presence as well as an increase in e-commerce shopping. Another recreational use of time would be for traveling. With more time available, without the stress of driving, we may see an increased willingness to travel in both short and long distance trips. Since riders do not need to keep their eyes on the roads, we see increased competition in video content for riders to consume during their ride and competition for marketing agencies to generate ads promoting goods and services [10]. Despite the growth in media and entertainment, the extra 46 minutes can also be used to finish up work-related topics during the commute to the office.

Trucking

According to the American Trucker Association, there are 3.5 million professional truck drivers in the U.S., and an additional 5.2 million people employed within the truck-driving industry. Introducing autonomous vehicles will greatly disrupt the trucking industry as 8.7 million trucking-related jobs may be replaced. Delivery companies will benefit in profits by eliminating wages; however, truck drivers will see unemployment.[1]

We are already seeing distruption in the trucking industry. A notable example would be Otto, an Uber-owner company. Otto is a self-driving technology company who manufactures hardware for trucks to allow for self-driving capabilities. Otto does not build the trucks themselves. Instead, Otto installs hardware such as radars, laser sensors and cameras onto pre-built trucks [2]. An example of a truck with Otto hardware installed can be seen driving on the highway to make its first beer delivery, 120 miles towards Colorado Springs. [3]

In addition, the transportation of goods provides no value to the goods being sold. Delivery from transportation drivers requires high labor costs, resulting in higher price of goods. By having autonomous trucks carry out deliveries, this will eliminate labor costs and ultimately lower the price of some goods.

Employment

Unemployment

Unskilled labor will be affected the most by the introduction of autonomous vehicles as they replace jobs for a lower cost. Jobs affected include drivers, police and traffic court administrators, automobile mechanics and others[4]. Job displacement will be a gradual effect; therefore, it is encouraged to acquire new skills to avoid being displaced by autonomous vehicles. Public transportation and delivery transportation that is not yet automated will see unemployment as autonomous vehicles come closer to reality. Without a job, displaced drivers will also put stress on the government to provide them with unemployment compensation until they acquire the new skill for another job. Not only does unemployment in public and delivery transportation affect the job security of the driver, but it also economically affects small towns and communities. Because no drivers will be passing through the small towns and communities, no drivers will eat in local restaurants or stay in the local motel. These small towns and communities may lose revenue due to less human traffic and result in the closure of small stores. The same is also expected on gas stations. Because of this chain reaction, it may cause a number of ghost towns to appear. [5]

Employment in New Industries

With job displacement, we will also see areas of growth in employment. Car media market, advertisement, and data analytic jobs are to name a few. Specifically, it is recommended to seek for start-up opportunities that complement future technology. This may include acquiring additional skills to become an autonomous vehicle mechanic or tap into areas such as multimedia for passengers to consume since they no longer need to be focused on driving.

Infrastructure and Housing

Parking

According to Donald Shoup, a Professor of urban planning at UCLA and an expert in the economics of parking, 95.8% of cars are left parked and left idle[6]. In Shoup's book, The High Cost of Free Parking, he studied extensively the price of free parking and how free parking consequentially results in traffic congestion, wasted time, and air pollution due to increased interest in parking in the area [7]. Introducing autonomous vehicles will reduce the number of human-driven cars and reduce the number of cars left idle. A common prediction is that there will be increased participation in ride-sharing which means self-driving cars will be constantly moving from one destination to another and will not require as much parking space[8]. In addition, self-driving cars can be parked close together since no one needs to get in or get out of the vehicle; therefore, saving, even more space to be used for other purposes such as housing that is affordable for a growing population.

The Suburban Compromise

Purchasing a place to live in the suburbs generally cost less per square footage compared to a place in city centers. The suburbs usually consists of more stand-alone residential houses that encourage privacy and space that allows for building a family. However, many jobs revolve around city centers, which are known to have a higher cost of living. Therefore, the suburban compromise represents the willingness to commute to and from work on a daily basis to fulfill the desires of personal space and nature [9]. Those who are commuting from the suburbs to city centers usually use a personal car, public transportation, or a mix of both methods to get themselves to and from work. Self-driving vehicles will open the doors of the suburbs to more people looking for stand-alone houses since transportation will be faster and they will not need to stress over driving. As a result, we should see an increase in population in the suburbs as self-driving vehicles start sharing the road.

City Centers

In city centers, we can generally access commerce, entertainment, and parts of history as well as politics within close proximity [10]. City centers consist of more commercial space rather than residential space and are generally associated with high living costs. As a result, those who work in areas within city centers tend to endure the suburban compromise. Introducing self-driving cars will give families and individuals more options in where they want to live since they would not need to spend the time driving to their destinations. Also, because more people will move outwards toward the suburbs due to desires to own larger properties, city centers would be less congested and the cost of living in city centers may decrease as a result of decreased housing demands.


An Introduction to Autonomous Vehicles

The Boring Company: Underground Freeways

The Boring Company, founded by Elon Musk, is inspired to solve the problem of Los Angeles traffic by proposing a multi-layered tunneling project that creates a massive underground freeway. Musk’s goal is to connect cars cross-country through a “fully autonomous journey across the US”[1]. The proposed idea will dramatically impact the infrastructure of cities and how we travel throughout the city or even country. The Boring Company has broke ground in February 2017 as they began digging from the SpaceX offices in Los Angeles [2]. The underground tunnels will be accessed by stopping the car on an electric sled that lowers the car to the underground freeway via a car elevator. In late July 2017, Musk posted on his Instagram a video of The Boring Company testing new elevators with a Tesla Model S [2]. The car elevators are similar to the elevators seen in the concept video shown by Musk at TED2017 in April 2017.

Green City (Fuel vs Electric)

Fuel

Postive

  • For personal cars and trucks, automation could save around 15 percent fuel through computer control which maintains a speed and accelerate brake control. According to JD’s power report “Why weight is so critical in Today’s Vehicles ”
  • “According to research from the Department of Energy (DOE), automated vehicles could reduce energy consumption in transportation by as much as 90%, or increase it by more than 200%. More than a quarter of greenhouse gas emissions come from the transportation sector, according to the Environmental Protection Agency (EPA).” The influence of the autonomous vehicle emission could be positive or negative but it will bring a huge change to the society.[3]

Negative

  • Automated cars can safely travel way faster than human-driven vehicles—computers have much quicker reactions than even the best human drivers—which matters because fuel economy typically decreases at speeds over 50 miles per hour.
  • Automated vehicle makes long distance travel more easily, people would like to ride their automated vehicle go anywhere they can go. That means the gas consumption will increase and will result in extra exhaust emission.[3]

Electric

Positive

  • Electric autonomous vehicles are easier to drive by computers control. Therefore, personal use or ride-hailing services will increasingly make up a higher percentage of daily miles driven. Electric autonomous vehicles will have zero emission compared to the fuel autonomous vehicle, whether the daily drive distance increase not.[4]
  • Autonomous vehicles enjoy a symbiosis with electric vehicles. The electric autonomous vehicle is used drive-by-wired, brake-by-wired and steering-by-wire rather than mechanical, driving and braking controls. Therefore, manufacturers can reduce the weight of the vehicle and produce more economical vehicles. This means lesser battery usage and longer driving distance. Meanwhile, purchasing and using the cost of those economical vehicles will be lower, this attracting customers to purchase the electronic vehicles rather than the fuel ones. It is another way to reduce emission [5]

Negative

  • All-electric vehicles (EVs) and plug-in hybrid electric vehicles (PHEVs) running only on electricity have zero tailpipe emissions, but emissions may be produced by the source of electrical power, such as natural gas, coal, and many more. Over 64% electric uses natural gas and coal, during this processing, it may cause more emission. [6]
Electric Source

Vision for the Future

The social and economical impacts of autonomous vehicles may be grim for workers with mundane jobs. In 20 to 30 years, we expect roads, laws, and regulations to be fully ready for all vehicles at level 5 autonomy.

Design of the Autonomous Vehicles

futurecar.
Future of Autonomous Vehicles.

The design of the autonomous vehicle will drastically change. Although autonomous vehicles can reduce the number of vehicle accidents, there is still a possibility that it may crash into objects, buildings, or other vehicles. With no steering wheels or braking systems with vehicles at level 5 autonomy, vehicle designers must think of where to place the airbags. If the vehicle seats can turn in any direction, airbags must be placed in all directions to ensure the safety of the passengers. We believe that seat belts will still be needed because there may be unforeseen events that can happen on the road. In addition, vehicles may not even need car signal lights, windshield wipers, or side mirrors in the future.

Future Use of Autonomous Vehicles

With vehicles at level 5 autonomy, individuals can enjoy leisure time in the vehicle as they will not have to pay attention to the road. People do not need to take their driver's test in order to get the license to drive a vehicle. Our group envisioned that autonomous vehicles can be equipped with a 360 degree screen where individuals can enjoy playing video games for longer road trips. We may also see that autonomous vehicles can be summoned and requests can be executed with the use of an Apple Watch or a smartphone app.

In the future, traffic lights or road signs may not be needed as autonomous vehicles may communicate with each other instead. Vehicles are able to coordinate with each other and communicate which one will go first at an intersection. We believe that people will start to shift towards purchasing autonomous vehicles because of the many benefits it offers. Individuals can enjoy an increase of leisure time and productivity. We may not even need to buy vehicles anymore and use a vehicle on a subscription basis. Without having to purchase vehicles, individuals can have more disposable income to purchase other things such as housing. Today, vehicles are left in parking lots or on the streets where it is not being used. With less vehicles in general, cities can narrow the roads and expand the space for more housing, parks, or even bike lanes.

With safer roads and less vehicle accidents, there will be less need for emergency personnels like firefighters, paramedics, and police officers.

Authors

Angela Au Carmen Chung Derek Guo Benny Jiang Wilson Tam
aaau@sfu.ca cca179@sfu.ca derekg@sfu.ca bennyj@sfu.ca wta22@sfu.ca
Beedie School of Business
Simon Fraser University
Burnaby, BC, Canada
Beedie School of Business
Simon Fraser University
Burnaby, BC, Canada
Beedie School of Business
Simon Fraser University
Burnaby, BC, Canada
Beedie School of Business
Simon Fraser University
Burnaby, BC, Canada
Beedie School of Business
Simon Fraser University
Burnaby, BC, Canada

References

  1. http://www.bbc.com/news/technology-39741094 Retrieved on July 21, 2017
  2. 2.0 2.1 https://en.wikipedia.org/wiki/The_Boring_Company Retrieved on July 21, 2017
  3. 3.0 3.1 http://time.com/4476614/self-driving-cars-environment/ Retrieved on August 1, 2017
  4. https://www.usatoday.com/story/money/cars/2016/09/19/why-most-self-driving-cars-electric/90614734/ Retrieved on August 1, 2017
  5. https://www.forbes.com/sites/jeffmcmahon/2017/04/17/big-fuel-savings-from-autonomous-vehicles/#4b50eb904390 Retrieved on August 1, 2017
  6. https://www.afdc.energy.gov/vehicles/electric_emissions.php Retrieved on August 1, 2017
Personal tools