Autonomous Vehicles Fall 2017

From New Media Business Blog

Jump to: navigation, search

Contents

Introduction

Introduction Video

Autonomous vehicles are defined as full-time performance by an automated driving system of all aspects of the dynamic driving task under all roadway and environmental conditions that can be managed by a human driver.

Autonomous vehicles may differ in the system designed, but follow a similar process[1]

  1. External hardware, namely sensors (including cameras, radars, laser beams etc.) attached to the vehicles will actively analyze the vehicle’s surrounding environment.
  2. Internal software process the inputs gathered from sensors to create an internal mapping of the surrounding environment.
  3. With the user’s choice of destination, the system software configures a suitable route to follow.
  4. Instructions are sent to the vehicle’s actuators controlling acceleration, braking and steering, and the vehicle begin moving towards the destination.

Levels

To pinpoint the degree of automation within an autonomous vehicle, levels are used. According to the Society of Automotive Engineers (SAE) International standard, autonomous vehicles can be categorized into five levels according to the different level of automation (disregarding level 0):

SAE Levels [2]


Level 0: No automation occurs in this level; the car is manually controlled by the driver in all driving tasks.

Level 1: The vehicle at this level can handle one driving task in an instance, but cannot perform more than one simultaneously. Tasks usually include accelerate, decelerate and steering.

Level 2: At this level, the vehicle can handle two driving tasks simultaneously. However, the major aspects of driving are still handled manually. Vehicles with assisting features such as cruise control is an example of level 2 automation.

Level 3: Starting on level three, the vehicle is considered autonomous. Most dynamic driving tasks can be handled by the vehicle. However, the reliability and stability at this point is still questionable. When the system is unable to detect external environment accurately or determine the best course of action to changing surroundings, immediate intervention is required by the driver.

Level 4: Vehicles at level four show dramatic improvement in terms of stability from level three. System failure will not require immediate intervention from the driver, and the dynamic driving tasks are handled smoother under normal conditions. However, vehicles will experience difficulties in engaging with extreme road and/or weather conditions such as heavy snowing, which will cause heavy obstruction to the sensor’s accuracy in determining surrounding environment for the vehicle.

Level 5: On this level, vehicles are truly autonomous. All aspects of the dynamic driving tasks are handled perfectly, and the performance will not be affected by any extreme conditions in comparison to level four. Driver presence is not necessary.

Timeline

The history of autonomous vehicles shows slow progression in the 1900s. There are significant advancement seen in the 2000s.

From 2004 and onward, the DARPA (Defense Advanced Research Projects Agency) Grand challenge was hosted in 2004, 2005 and 2007[3]. The challenge was a prize competition, where the DAPRA (a research organization under US Department of Defense) reward highly innovative research to materialize autonomous vehicles.

In the 2004 challenge, no teams were able to complete the 240km route in a desert setting. However, in 2005, five of the teams entered managed to finish the 212km route in a similar setting as the previous challenge. The winner of the challenge is Stanford University’s “Stanley”, that is developed by their very own Stanford Racing Team lead by Sebastian Thrun, who was the director of Stanford Artificial Intelligence Lab at the time. The victory rewarded Stanford with two million in prize. In 2007, the challenge was held in an urban environment, and won by “Boss”, a modified Chevrolet Tahoe by Carnegie Mellon University’s Tartan Racing[4].

Thrun’s contribution to autonomous vehicles became widely known after Google’s announced their Self-Driving Car Project which is led by Thrun and other leading experts in the area in 2010[5].

By 2013, major car manufacturers including BMW, Volvo, GM, Ford, Mercedes Benz, Audi, Volkswagen, Toyota and Nissan have made certain investments in developing driverless car systems. At this period, features provided were not fully autonomous, and only provided assistance to the driver, such as adaptive cruise control[6].

In 2014, Tesla announced development on autopilot feature for the Model S, and by October 2015, version 7 of the internal software in the vehicle had incorporated autopilot feature. The feature was available to owners of the Model S via an overnight update[7].Following this announcement, in May 2016, a fatal crash happened during the testing of Tesla’s autopilot function in a Model S, ended up killing the driver. The crash occurred when the vehicle was driving down a divided highway when a white semi-truck cut across the highway, and under the brightly lit sky, the system failed to identify the truck. This resulted in the vehicle continuing to accelerate without applying brake, and ran directly under the truck, causing a fatal crash[8].

In the same year, Singapore announced the first ever autonomous taxi service (car hailing) supported by nuTonomy[9].

By 2017, Audi has announced their A8 model will have system capable to carry out level three autonomous driving[10]. Meanwhile most car manufacturers embraced the idea to skip level three development and head straight for fully autonomous functionality. This is due to the level three’s nature in requiring immediate intervention from driver when the system fail, and it presents significant risks to the drivers[11].

In the same year, Elon Musk, CEO of Tesla Inc, announced the timeline for conducting cross-country test of the proposed fully autonomous Model S from Los Angeles to New York by the end of the year[12].

In November 2017, Waymo began testing self-driving cars on the road in Phoenix, Arizona. The ride-hailing service is offered to local residents for free in encouraging adoption of autonomous vehicles. Although the service was already offered earlier in April, there was a safety driver behind the wheel to take over when the system fails. Now the testing will consist of no human drivers to intervene. Waymo is also looking into autonomous trucks and consider selling their technologies directly to automakers. [13]

In the near future, Delphi and Mobileye proposed to introduce a level four self-driving sytem by 2019[14].


Examples

Tesla

Tesla Autopilot 2.0 Full Self Driving Hardware - Neighborhood Long

Tesla is in the implementation stage of its autonomous driving features. Unlike other companies, it doesn’t rely on LiDAR technology. Instead, 8 surround cameras provide 360 degrees of visibility around the car at up to 250 meters of range, and 12 updated ultrasonic sensors complement this vision allowing for detection of both hard and soft objects at nearly twice the distance of the prior system. The system is also designed to be able to conduct short and long distance trips with no action required by the person in the driver’s seat [1].

With the summon feature, all you will need to do is get in and tell your car where to go. If you don’t say anything, the car will look at your calendar and take you there as the assumed destination or just home if nothing is on the calendar. When you arrive at your destination, simply step out at the entrance and your car will enter park seek mode, automatically search for a spot and park itself [2].

 

Google Waymo

Waymo is giving out test rides in self-driving cars

The Google self-driving car project began since 2009. In 2015, “Firefly” hit public roads for the first time.

However, recently, Waymo is retiring its autonomous “Firefly” vehicles. Waymo will now focus on integrating its self-driving technology into the Chrysler Pacifica minivan in an effort to bring the technology to the mass market [1].

Waymo announced the Early Rider program earlier this year that it would give hundreds of Phoenix residents free rides in its self-driving Lexus SUVs and Chrysler Pacifica minivans. The goal of the program is to give participants access to their fleet every day, from work, to school, to the movies and more. This program gives Waymo a chance to learn about how passengers will react to the cars, and what type of experience is most pleasing [2].


Renault

Renault SYMBIOZ: concept and vision for mobility

In September 2017, Renault has revealed an autonomous, electric concept car that it sees as an extension of the home – the company describes the vehicle as "an extra mobile, modular and multi-purpose room".

The car's exterior is designed to resemble the architecture of the contemporary home, with a steel frame, expansive windows and wooden joinery details. The vehicle also connects to the home wirelessly, just like many of today's domestic appliances. Inside, the living-room-inspired interior includes four seats that can pivot to face each other. The design also includes a platform that allows the car to be lifted up onto a roof terrace [1].

Although Renault doesn’t go into detail about the connectivity and entertainment options, it does claim that a real-life demonstration will be available for testing later this year.


 

Smart

Smart Vision EQ Fortwo

The smart Vision EQ Fortwo is the first Daimler concept without a steering wheel and it’s a fully autonomous car, which is set to go into production by 2022. It provides a new vision of urban mobility and highly flexible local public transport: it picks up its passengers directly from their chosen location. It uses a rechargeable 30 kWh lithium-ion battery. When not ferrying passengers around, it drives to a charging station to replenish the battery all by itself [1].

In the future, users will be able to order and operate the vehicle - which will drive at Level 5, the highest level of automation - entirely through their phones[2].

The smart CEO says “The smart vision EQ fortwo is our vision of future urban mobility; it is the most radical car sharing concept car of all: fully autonomous, with maximum communication capabilities, friendly, comprehensively personalisable and, of course, electric.” [3]


Benefits

Fuel Savings

Autonomous vehicles can create significant fuel savings and reduce fuel consumption by as much as 44 percent for passenger vehicles and 18 percent for trucks.[4] With the car's driving and braking controls being employed wired rather than mechanically, manufacturers can eliminate the weight contributed by those mechanism.[4] Ultimately, this produces a more economical car. Similarly, a much smaller and lighter vehicle is produced for a single-passenger office commute.

With so much fuel waste comes from inefficient driving itself such as gunning it at green lights or slamming on the brakes, autonomous cars can help eliminate this. With the cars being built to optimize efficiency in acceleration, braking, and speed variation, they will help increase fuel efficiency and reduce carbon emissions.[5]


Economic Benefits

Self Driving Car Test: Steve Mahan

Many individuals will personally benefit from self driving cars as they enable the elderly to be mobile as well as empower those who are blind or have a disabilities. Even individuals who are unlicensed or who do not own a car will be able to travel as well. This allows many who have been unable to drive in the past, to become more independent and be in control of themselves. [1]

In 2012, Google showed how powerful this technology could be by helping empower people to become more independent. This was shown with Google assisting a blind man while he ran errands in one of their self-driving cars. Steve Mahan who is 95% blind was able to perform his daily errands by travelling in Google’s modified Toyota Prius. [2] This is just one of the many examples of autonomous cars helping those who are unable to drive themselves.

Also, the adoption of driver-less fleets will mean you will never have to search for a parking space again. This is because you will always get dropped off at your desired location. However, If you do opt to own your own autonomous vehicle, there will still be no need for you to park it because it will locate a space and park itself. [1]

 

Traffic Efficiency

Platooning [3]

Autonomous cars will also decrease the amount of traffic congestion. Most time spent in traffic is wasted time, but Autonomous cars will enable drivers to spend that time doing something more productive instead such as reading a book, watching TV or catching up on work. As a result, drivers in autonomous vehicles will not have to give their full, constant attention to driving and can shift their focus to other tasks. More would be accomplished in a shorter period of time.

A concept called truck platooning can also reduce highway congestion. Platooning is when autonomous trucks and cars are grouped together using electrical coupling, enabling simultaneous braking and acceleration.[4] This will enable cars to avoids the “accordion” effect of vehicles speeding up or slowing down at unpredictable times. With autonomous cars having high-tech sensors that can react 90% faster than humans, platooning allows the distance between trucks to fall from 50 meters to 15 meters.[4]

Autonomous cars will also be able to follow the best routes available according to real-time information. These cars may be able to read the condition of the roads in real-time and redirect the routes accordingly. Similarly, vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) are developing technologies that enable vehicles to exchange safety and mobility information with one another and surrounding infrastructure.[4] Vehicles can receive information like road conditions while en-route and shift accordingly or autonomously coordinate behaviour with other vehicles and infrastructure like at intersections.[4]

 

 

Safety

About 1.24 million people worldwide die each year in collisions, and that number is expected to skyrocket to 2.2 million by 2030, according to the World Health Organization. However, autonomous cars have the ability to dramatically reduce the number of accidents, thus potentially saving millions of lives.[5]

Similarly, according to the National Highway Traffic Safety Administration (NHTSA), 94% of accidents are related to human error. Automated vehicles can remove human error and prevent a huge majority of accidents.[4] With the concept of platooning, there would be a 500% increase in greater lane capacity, vehicles per lane per hour. Lane capacity can also increase with the adoption of adaptive cruise control. Adaptive cruise control automatically adjusts vehicle speed to maintain a safe distance between cars and trucks ahead. This allows drivers to stay on cruise control for miles and avoid having to speed up or slow down depending on traffic.[4]


Disadvantages

Concerns With Car Hacking

Jeep Example: Hackers Remotely Control a Jeep

Other than accidents caused by computer malfunctioning and glitching, car hacking is deemed the biggest fear with self-driving vehicles. Car hacking is essentially losing control of the vehicle due to a hacker taking remote control of the car. This could include acceleration and sudden braking. There have been no reports to date of malicious attackers targeting driverless vehicles. The widely known Tesla and Jeep hacks were executed by cybersecurity professionals who had a substantial amount of time to work with the cars to demonstrate their skills for hiring purposes. [1]

One thing to note is that Tesla owners will someday have the ability to summon their cars remotely using their mobile app [2]. The app is similar to Apple’s Find My iPhone, where real-time location data is sent to owners’ apps. This could cause concerns in phones being compromised, and thus, having access to the vehicle.

 

 

Won't Decrease Traffic

Another factor to consider is that autonomous vehicles may not actually reduce traffic congestion. According to Lew Fulton, co-director of UC Davis’ Institute of Transportation Studies (ITP), autonomous ride-sharing could become so cheap that there would be no incentive to carpool, putting more cars on the road. [3]

Elon Musk proposed allowing Tesla owners to rent out their autonomous cars to rideshare users. However, this could lead to hordes of autonomous “zombie cars” driving around with no occupants inside between picking up passengers, which will create congestion as a result.

One way to try and combat congestion is to institute a fee that discourages zero-occupancy vehicles. A law that was proposed in Massachusetts suggested a fee of at least $0.025 per mile to be enacted [4].

Ultimately, it all comes down to how people will use driverless cars; if the increased convenience leads to greater usage, then the benefits of traffic efficiency may not be reaped. If people can sleep or get work done in their cars, they may be more inclined to move farther out of city centers and depend on cars to get around, according to Fulton [5].

Driving Ability Dependent on Weather

Ford testing its autonomous vehicle in winter weather (Michigan).

In order to see and operate, autonomous cars rely on cameras and sensors to build a computerized view of the car’s surroundings. This becomes more difficult in adverse weather conditions when visibility is reduced. Ford has tested its autonomous vehicles on very light snow in Michigan, using Lidar sensors and comparing the information to a high resolution map of the road stored in its onboard computer [1]. In order to have these high-resolution maps in its system, the car has to have driven the route before – this is one of its limitations. Travelling along a new route in the midst of a snowstorm could be problematic.

As Ford was merely testing on very light snow, the snowstorms of a Canadian winter evidently poses bigger challenges.

 

 

 

Ethical Dilemma: The ‘Driverless Dilemma’

The Trolley Problem

Autonomous cars will have a massive effect on society, as there are entire industries built around cars. For example, car insurance, parking industries, and long-haul truck driver jobs. Furthermore, if there are no more truck drivers, there will be a reduced need for rest stops, food shops, and hotels along the route.

However, with driverless cars, a new ethical dilemma is raised: The Trolley Problem. The Trolley Problem is becoming real; the difference is, the self-driving car is analogous to the train headed towards pedestrians on the road. The only way to save these pedestrians is to swerve out of the way, but that will run the car into a concrete wall, and will kill the passenger inside the vehicle.

The implication of the Trolley Problem is that most people would not buy an autonomous vehicle if it meant that they had to sacrifice themselves to save the lives of others.

 

 

Implications of Autonomous Vehicles

Regulations

Current Laws

The regulations have to be updated to accommodate the introduction of autonomous vehicles; currently most of the states in U.S. have introduced laws regarding autonomous vehicles and it varies across different states.

State Map

Here is a map of the latest U.S states where it displays the current status of autonomous vehicle regulations in different states

Currently there are 21 states and Washington, DC have enacted bills related to autonomous vehicles[1]. Among them Nevada, California, Tennessee, Michigan, Florida, and DC (Washington) are the pioneer and they are the one that likely to set the standards for others[2].

California, a popular state for the testing of autonomous technology with 42 companies hold permits recently revised its current regulation.The new regulations will allow the testing of autonomous vehicle without driver behind the wheel in 2018, and allow the public to use the autonomous vehicle as long as they are not charged. It also requires the manufacturers to notify the local government about their testing plan, updates of the technology and to report the number of times that human intervention required as the vehicle fail to navigate.The testing of autonomous truck remain prohibited in California[3].

Other states are taking actions to enact similar laws to attract investment and remain competitive.Texas pass the Bill 2205 that allow the testing of autonomous vehicle on states road and highway without driver behind wheel[4] . Ohio is spending $218 million building a 56 km long autonomous car highway on Route 33 to attract companies to test their autonomous technology in extreme weather. The highway will be covered with fiber optic cables to allow the sensor to communicate traffic, road conditions and the change of weather to autonomous car via Wifi network. The cables and sensor will also connect with government vehicles using short range radio transmitters.The highway is expect to be complete in 2018[5].


The SELF DRIVE (Safely Ensuring Lives Future Deployment and Research In Vehicle Evolution) act is increasing the federal motor vehicle safety standard (FMVSS) exemption caps from 2,500 to 100,000, meaning it will allow automakers and tech firms to test and eventually deploy 100,000 of autonomous vehicles without meeting local and states safety standards[6] However, the public is concerned with the safety of autonomous vehicle. The SELF-DRIVE Act will require[7]:

  • NHTSA (National Highway Traffic Safety Administration) to enforce/enact a rule requiring manufacturers to submit a safety assessment certification for highly automated vehicles and automated driving systems.
  • Manufacturers to submit safety assessment letters described in the Federal Automated Vehicles Policy (or any successor guidance).
  • Manufacturers to develop detailed cybersecurity plans for automated vehicles.

The Act also clarified the role of NHTSA in regulating the development of autonomous vehicle. An advisory council will be set up to assist NHTSA to keep up to date in the field.

 

Data Privacy

Data will become the most valuable asset in autonomous vehicle, and everyone including consumer, insurance, regulator and manufacturer need to have this information[8].

Here are the types of data collected in autonomous vehicle[8]:

  1. Driving – Information generated by vehicle. For instance, speed, time, date of travel, acceleration and breaking pattern. The information is valuable to underwriting efforts and liability issue for carriers.
  2. Activity – information collected from passenger during the ride. For instance, identity, demographic of passenger, conversation and activity in the car which protect insurer from liability claims if recorded audio (data privacy & ownership) permitted
  3. Environmental – All the information collected outside of the car. For instance, road condition, location, and other car movements. It is hard to quantify the value of these data, but it allows insurer to evaluate potential risk more precisely.

As more data are collected, privacy implication and the potential misuse of this sensitive information is a concern to consumer. In a breach, consumer might suffer losses. Thus, protecting this information is key to gain customer trust and enable the growth of the platform. The SELF-DRIVE act will require manufacturers to develop a written privacy plan which specify how data will be collected, used, shared and stored, while clarifying that the Federal Trade Commission will continue to enforce violations of these requirements under Section 5 of the Federal Trade Commission Act[7].

 

Other Areas of Concern

There are several other areas of the current regulations that require further studies, otherwise they will act as a barrier to the adoption rate of autonomous vehicle. Australia’s National Transport Commission (NTC) is currently studying the following aspects in order to have a clear set of regulation by 2020[9].

Drinking and Driving – Currently, it is illegal for a drunk person to manually drive a car. It is argued by the NTC that this should not apply to a fully autonomous vehicle since the person is not in charge of driving at all, the drunk person only provides a destination to the system. However, the current law is still applicable to cars that fall under level 3 or below. Level 3 cars are considered autonomous but still require certain human interventions. Thus, it will be difficult for traffic police to determine whether the car is in manual or autonomous mode.

Duty and Liability Issue in Car Accident - The liability issue in a car accident is clear when it is a fully autonomous car or a manual car. In a manual car, the driver is responsible for his/her contribution to the accident. While in a fully autonomous car, the product itself will likely be responsible for the accident. The line is blurred when the car is semi-autonomous, as it will need to determine the car’s driving mode versus the driver’s responsibility at the same time. In a fully autonomous car accident, the car will remain at site and report the incident by itself or may require the rider/owner to report the incident. What happen when no one is in the car?

Parking Fee – Who is responsible to pay for parking fee when there is no passenger in the vehicle, currently the owner of the car will be penalized if no one pays the parking fee.

Seat Belt – For safety purpose and compliance, the driver is responsible to make sure passenger under certain age wear a seat belt. With autonomous vehicle, the owner will be held responsible if passengers did not wear a seat belt when the owner is not in the same ride or the car will not start unless the passenger wears a seat belt.

Take Direction from Law Enforcement – The autonomous car will need to have ability to take instruction from such situation.

Portable Warning Sign – Who is responsible for putting the portable warning sign out when an autonomous car breaks down or autonomous truck stops. Is the car going to send out signal to other cars nearby or the passenger have to put the warning sign out?

 

Economy

Insurance Company

The primary impact of autonomous vehicle is the sharp decline of revenue. According to a report by KPMG in June 2017, the size of automobile insurance sector is estimated to shrink by 71% ($137 billion) by 2050. Base on the survey conducted by KPMG in June 2015, where the respondents are executives from the insurance companies make up about 85 billion in personal and commercial premium[8].

Here is some of the survey result and it is clear that the insurance industry is not ready for such movement[10]:

  1. 45% of the respondents indicated that their personal auto premium charged will decrease, and 35% indicate that commercial auto premium will decrease due to the emerging autonomous technology.
  2. 74% of the respondents indicate that their company are not ready for autonomous vehicle.
  3. 65% of the respondent indicate they only have internal/external conversation. 10% indicate they have develop stratgeic plan to deal with autonomous vehicle.
  4. 58% indicate the manufacturer will be a major player in auto insurance in future, while 45% indicate it will be the new start up business.
  5. 68% of the respondent’s company did not allocate any budget to prepare for the change, and only 3% of them allocate >1% of their operation budget on it.
  6. 81% of the respondent indicate that established tech firm will own the autonomous driving data, while 71% indicate the manufacturer will own the data.

As for today, some insurance company are still unprepared. Some of them are misled by the increase in claim frequency from 2015 till today due to the better economy which fuel more driving and more distracted drivers on road[8].

The sharp decline in revenue is primarily due to the following factors[8]:

  1. Decline in Accident Rate and the Serverity of Car Accident
  2. The Growth of Sharing Economy
  3. New Entrants and Stiff Competition

Decline in Accident Rate and Serverity of Car Accidet

According to National Highway Trafic Safety, 94% of car crashes in US in 2015 are associated with bad decision made or human error [11]. With the adoption of autonomous vehicle, it is estimated that 90% of the car accidents will be elimated. Currently, the car premium insurance rate is caculate base on factors like car model, location, deductible, claim history. With significant lower car accident rates, there is reudction in loss – the amount of money that insurer has to paid to cover claim plus cost of adminstering and investigation the claim. Thus, the insurance company have to reduce their premium to compensate the reduction[8].

The cost of future claim is less uncertain given that more expensive technology components are used in autonomous vehicle. For instance, the cameras and sensors used. But given the economic of scale effect, the cost of production in some autonomous vehicle component will be drive down. For example, the Light Detection and Ranging sensor (LiDAR) by Velodyne is able to drive down their price significantly from $80,000 in earlier days to $8000 now (drop by 10 times) with investment from other sources. Velodyne is aiming to have their Lidar sensor to cost in the mid to lower 100s by end of 2018[8].

With autonomous techology, cars can identify & react to an imminent accident more quickly which minimizes the impact. Accident happen at lower speed is much less severe than those at high speed[8].

Growth of Sharing Economy

For the past few year, the mobility on demand platform became a common option for many urban and suburban residents due to the convenience and cost saving benefits. The rid hailing industry is estimate to reach $285 billion by 2030 [12].

The move towards shared vehicle rather than ownership will have profound impact on auto insurance. Consumers are ready to adopt a new transportation model given that the number of connected vehicle used for ride hailing tripled over last 2 year. It is estimated there will be more than 1.5 million of connected car by end of 2017[8]. Another study indicate that personal car ownership in United State will drop by 80% by 2030 (from 247 million in 2020 to 44million in 2030) and 95% of the passenger miles will be served by an on-demand autonomous vehicle. [13].

Insurance Change in Landscape[8]

As mentioned earlier, the US auto insurance will likely experience a 71% shrinkage by 2050. The landscape of auto insuance also changes due to more commerical vehicles on road and product liability; a new type of liability emerging from autonomous technology. Although there is no precedent case or any legal postion defining product liability yet, automanfacturer like Volvo, Mercedes and Google announced that they will accept reponsiblityand liability for any accident caused by their future autonomous vehicles[8]. From the graph, it is clear that product liability will account for 57% of auto insurance in future, and the current dominant personal auto insurance will shrink from 86% to 22%.

New Entrants

With the adoption of autonomous vehicle, the current insurance sector will face the threat of new entrants. Manufacturer or technology firms who develop autonomous vehicles may enter into the insurance market to get a slice of the pie.

Manufacturers run their own insurance company[8]

Currently, we buy insurance from an agent or insurance company whom captures point in time driving data and underwrite the policy on periodic basis. There is no direct connection between consumer and manfacturer in buying insurance.

There is no regulation regarding autonomous data ownership, even recent SELF DRIVE act did not answer the question when they require manfactuer to develop a privacy plan[14].We expect the issue to be addressed soon before autonomous vehicles become mainstream.

Under the assumption that the manfactuer are the only one that has access to the data by contract upon the purchase of vehicle. They can disintermeidate the insurance company as the data captured by the vehicle will allow the firms to gain insights of autonomy platform and driver performance. The firms can price individual insurance more precisely and at the same time enhance customer relationship. The insurance can be incoprated as either part of monthly payment or part of the retail pirce, hence, driving out the traditional insurance company[8].

 

Auto Manufacturer

With autonomous vehicle, the revenue of manufacturers will be eroded significantly.

Here are the main factors that led to the sharp decline in their revenue:

  1. Drop in After Market Part Sales
  2. 40% reduction in New Car Sales
  3. Increase in Product Liability

Drop in After Market Part Sales

With a 90% reduction in car accident rates, the after market part sales and services by manufacturers are affected significantly[8]. The after market part sales roughly account for 24% of their business revenue stream. Back in 2001, General Motor generates more profit from a $9 billion aftermarket sales compared to a $150 billion car sales[15].

Reduction in New Car Sales

Research proposed that 95% of the time a personal car remains unused; therefore, consumers will be more motivated to use the ride sharing platform rather than car ownership in future. It is estimated 40% of new car sales will be eroded due to autonomous vehicles [16]. Combined with the information earlier where there is an estimate that private car ownership will drop by 203 million from 2020 to 2030. [13]. There is a clear sign that the automakers will have a sharp decline in their revenue from car sales.

Increase in Product Liability

With decisions made by autonomous vehicles, car automakers will face increasing driving risks and liabilities.

 

Sourcing New Revenue Stream

Enter Mobility On Demand Market

The automanfactuer will likely enter the mobility on-demand and ride hailing industry to gain some market share given that the sector will continue to grow. In fact, some of the automanufacturer are planning to enter or are already in the market.

BMW ReachNow Concept

  • Volvo and Uber join partnership in developing autonomous vehicle
  • Ford’s mobility service within next 5 year
  • GM investment in Lyft
  • ReachNow and DriveNow by BMW


BMW runs its DriveNow program in Europe and ReachNow program in US. The ReachNow program is a hybrid of ride hailing and car sharing platform.

It uses a hybrid of models that is kind of distinctive from the others [1]:

  • Ride - it is similar to the ride sharing concept where customers get picked up by a trained professional driver similar to Uber.
  • Reserve model - this allow customers to reserve car ahead of time and have it deliver to them (where & when) and it will be collected when customer is done with it.
  • Drive model - Customers can drive any BMW or Mini model themselves for an extended period, at a rate of 41 cents/minute or $80/day. Fuel or charging, insurance and parking are included in the price within a "home area". Customer simply pick up available car nearby and drop it anywhere.

Enter Insurance Market

Alternatively,the automanfacturer can choose to enter the insurance sector. As driving statistics are captured by the autonomous vehicle, a wealth of information are collected which encompass the entire driving experience including speed, road condition, weather and other car movement. Even with a level 3 autonomous car, the behavior data is recorded which allow better understanding and precise measurement of driver’s relative risk. As mention earlier, the car owner will likely allow manfacturers to access the data via trasparent contractual agreements, financial incentives and for safety purpose[2].

The data will allow automakers to continue enhancing their autonomous vehicle’s algorithms which in turn improves the vehicles’s awareness of current road conditions. Furthermore, with the information collected, it may be profitable for automakers to enter into the insurance sector.[2].

Here is a hypothetical scenario to illustrate the extreme case that automanufacturers become insurance providers by themselves[2]:

Reduction in Premium [2]

The left side of the underwriting analysis picture shows the current auto insurance model where the $500 policy charged is the breakeven point. With premium revenue equals to the amount paid for commissions, policy management, tax and losses (including expense of handling claim).Currently, claims/losses account for 80% of the total cost, $345 on loss payment and $45 on handling claims[2].

With the new proposed insurance model, the claim cost could fall significantly due to the reduction of car accident and more precise measurement of premium base on the data captured. The claim cost could reduce by $116. If the automanufacturer run its own insurance company or incorporate the insurance to monthly payment/retail price, the commission payment can also drop from 15% to 5% as the dealer will require less work to get buyer committed to insurance. To incentivize car owners and enhance customer relationship, the automaker can offer about half of the cost($82) save to the car owner in a form of premium reduction, while still keeping $84 profit from a $500 premium policy[2].

Of course this scenario is made under the assumption that the data is kept to themselves and they lock in the insurance arrangement upon purchase.

 

Auto Repair Shops

Given that 90% of car accident is caused by human, autonomous vehicles will eliminate most of the car crashes. As a result, there will be a huge portion of losses for auto repair shops. The loss of revenue will be shared amongst the 228,600 repair and maintenance shops[3]. In 2010, property damages by 24 million vehicles amount to $76 billion[4].

 

Parkings

Autonomous vehicle will drop passengers off at their destination and park elsewhere or simply remain on the road as part of the ride sharing program. Currently US has about 144 billion square feet of parking spaces. It is estimated that autonomous vehicles have the potential to reduce parking space by about 61 billion square feet. Those space can be used for more social and economic purposes[4]. With more space available in urban area, the real estate price will be affected.

However, it also means significantly less money for cities through parking tickets. New York city has $565 million in parking ticket in 2015. It account for 59% of total fine and had experienced an increase of 3.5% from previous year [5].

 

Law Enforcement

There will be significantly less traffic violation such as speeding when autonomous vehicle becomes a mainstrem. This will reduce government revenue. The lost in government revenue may lead to a downsize of law enforcement forces[4].

Here are some traffic violation fines in NYC 2015[5]:

  • $29 million on red light camera
  • $17 million on bus lane
  • $31 million on speed camera

With less traffic violation, the law enforcement can shift focus on more important areas such as drug trafficking.

 

Media and Entertainment

Once the autonomous vehicle is fully integrated into the city, car drive to and from a designated location with a command.

On average, US commuter spend 50 minutes travelling on road daily.With the 50 minutes saved by autonomous vehicle, passenger can allocate the time to other activities. The time can be used to do work related activities while on the way to office or surfing on the internet. Each extra minute spent by vehicle occupants can generate $5.6 billion annually[6]. There will be an increasing competition in video content for riders to consume during their ride and competition for marketing agencies to generate ads promoting goods and services[7].

 

Transportation

Otto's Truck delivers Budweiser

Trucking

For transportation, the implications are rather vast. Trucking industry amongst the transportation sector will be one that is directly impacted by the introduction of autonomous vehicles.

For the trucking industry, labour costs play a significant portion for total cost of transporting goods. Truck transportation relies heavily on the drivers to deliver cargos safely to a destination. There are strong economic incentives to make trucking fleets early adopters of autonomous technology, where gradual progression had been seen on safer, cleaner and more affordable autonomous vehicles, especially in semi-trucks. With adoption of autonomous fleets, trucking industry can dramatically reduce the number of drivers required for each transport. Not only are the autonomous trucks more efficient in terms of fuel savings from optimizing control of the truck versus human drivers that may sometimes not be efficient in fuel usage; machines do not require any pit stops for rests. This would improve trucking efficiency not only in fuel savings, but also time per trip for transportation. This amount to significant cost savings for the companies.

On the other hand, with trucking companies’ eagerness to adopt autonomous trucks, there will most likely be loss of employment when drivers are replaced gradually by the machines. A trend that is predicted to occur is the shift for drivers to take on jobs with more analytical aspects and skills requirement. This would mean initiating a transition process for drivers to learn new skillsets. Another aspect that may be impacted from autonomous vehicles relevant to trucking is the loss of businesses for small towns. This is a result from reducing and potentially eliminating human drivers, and consequently leading to reduced traffic into smaller towns that may depend on travelling long distance truck drivers to make pit stops previously. Towns may struggle to survive potentially.

In terms of the progress for introducing autonomous vehicles into the trucking industry, there are some breakthroughs. In April 2016, 12 Caravan truck finished a trip across part of Europe aiming to utilize platooning technology, which would allow commercial trucks to follow one another closely to reduce drag, and at the same time improve safety and potential to create economic growth in traffic and transportation sector. [1]

In the same year, Otto, a startup firm acquired by Uber partnered with Budweiser to pilot an autonomous truck to drive through Colorado in order to deliver a cargo of beer over a total distance of 120 miles without any human intervention. The delivery was a success and symbolizes a huge advancement for autonomous trucks to become reality in the near future. [2] The video from WIRED describe further on the journey that the truck took.


Aircraft

Airbus's Vahana at the Geneva Motor Show

For aircraft industry, there are significant advancements in this sector as well in recent years. Although auto-pilot already exists in many of the commercial airplanes, companies are considering taking the next step to convert airplanes into completely autonomous aircrafts. A leading innovator in the industry is Boeing. They have announced in 2017 that they are developing a self-flying airplane, and will begin testing soon in the near future [1]. They have solidified their interests in advancing autonomous initiative by announcing to acquire Aurora Flight Science in October 2017 [2]. The company is a startup that had been developing its Lightning Strike XV-24A vertical take-off and landing craft in recent years. The aircraft is an autonomous military aircraft that is funded by DARPA and the USAF. Earlier this year, Aurora had ran a successful test flight program for its SV-24A Demonstrator, a subscale version of the eventual production vehicle.[3].

Another company that is taking the imitative to advance the progress for autonomous aircrafts to become reality is Airbus. They unveiled their concept at the Geneva Motor Show in 2017, in collaboration with Italdesign to introduce a flying car named Vahana. Airbus announce testing will begin before the end of 2018.[4] The video on the right shows the concept illustrated by Airbus and Italdesign on the futuristic flying car as a taxi service for people.


Authors

Lauren Chan Sally Li Wing Liang Haofan (Zack) Lin Jean-Ray Tseng
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


Reference

  1. http://www.independent.co.uk/life-style/gadgets-and-tech/news/boeing-start-testing-self-flying-planes-autonomous-aircraft-no-pilot-artificial-intelligence-a7798901.html Retrieved on November 16, 2017
  2. https://techcrunch.com/2017/10/05/boeing-to-acquire-aurora-flight-sciences-in-bet-on-autonomous-flight/ Retrieved on November 16, 2017
  3. https://techcrunch.com/2017/04/04/auroras-massive-lightningstrike-vtol-uav-just-got-one-step-closer-to-reality/ Retrieved on November 16, 2017
  4. https://techcrunch.com/2017/10/05/airbus-on-track-to-fly-its-electric-aerial-taxi-in-2018/
Personal tools