Consulting Report APA style with 8 pages (3750 min words)



Autonomous Vehicles: A Complex ESEM System

Date Submitted: November 14, 2017

Total Score = 100

Criteria Description Grade % Score Comments

Good content and analysis

with specific recommendations

to their client. 25% 25

Very well done in all respects


Paper has introductory,

evidential, and conclusive


25% 25


Word Count Stays close to

range of 3,750 to 7,250 words.

25% 25 4039

Good grammar

It is not hard to interpret the meaning of

statements because of poor grammar.

12.50% 12.5

Very well written

Comment [ENB1]:


Proper References

Paper has both APA format in-text and bibliographic

citations (numbering at least 5).

12.50% 12.5

Right on target

CEE 400 Earth Systems Engineering & Management

Brad Allenby & Bruce Marsh Fall 2017

Table of Contents

Executive Summary...............................................................................................................(3)








Executive Summary

Autonomous vehicles, as a recent and constantly growing complex system, are likely to

take over the automotive industry. Before full commercialization can occur, many of the

technical and socioeconomic implications need to be assessed. These are analyzed

through the scope of ESEM governance and theoretical principles and mainly from the

perspective of the United States. It was determined that mitigating risk due to cyber-


attacks, reducing carbon emissions, and having an effective software system should be

the focus going forward.

As such, based on the conclusions drawn in this report, the recommendations for Tesla, Inc. are as follows:

1. Include at least (2) forms of failsafe that can be used in the case of a malfunction or targeted attack on the system, or any other potential complication. (i) An emergency manual override switch that turns off all automated driving

systems to allow it to function like a normal vehicle. (ii) The ability for the software to either detect an intrusion, or detect that

something is not functioning as it should and to notify the passengers.

2. Ensure that the autonomous vehicles are environmentally friendly and satisfy the community and other Tesla stakeholders. (i) Carbon emissions can be minimized by implementing completely electric


3. Communicate with legislators to ensure legality and safety of autonomous vehicle systems in the given jurisdiction.

4. Emphasize research and development of software systems, as they will soon become a larger portion of the automotive market. (i) Invest in creating a robust and effective proprietary software with an elevated level of cyber security.


Self-driving cars, or autonomous vehicles, are vehicles that can operate and navigate

on their own without a designated driver. Autonomous vehicles can be separated into

four categories depending on the level of human intervention, ranging from full human

control to none. Recent models can function without any human intervention

whatsoever, relying entirely on their sensors to detect the environment. This report will

focus primarily on the fourth category in which the vehicles operate on their own while


carrying passengers. Although they are a fairly recent technological advancement,

autonomous vehicles have been a topic of discussion in media and popular culture

since automobiles became commonly used nearly a century ago.

The early 1990s is when development started to take off after the United States

Congress passed the ISTEA Transportation Authorization bill. This bill pushed the US

Department of Transportation to demonstrate a fully functioning automated vehicle and

highway system, a task in which the largest companies and organizations at the time

undertook. Some of the first truly autonomous vehicles were developed during this

period, such as the VaMP and the VITA-2 which are the products of eight years of work

and millions of dollars in research and development. These vehicles were able to drive

long distances in standard highway traffic and speeds without any issues. As more

companies started to invest in autonomous vehicles the industry grew larger and

legislation quickly followed. Many states began enacting laws specifically for the

operation of such vehicles on public roads with Nevada being the first state to allow it in

2011. The regulations became progressively lenient and as the public became more

comfortable with the idea, general opinion shifted towards support of these laws.

This report will examine and assess the technological, environmental, and social

aspects of autonomous vehicles through the scope of Earths Systems Engineering

Management (ESEM) principles. These principles are divided into two major categories,

theoretical and governance. The theoretical principles are the core ideas behind ESEM

and they outline how to effectively interact with complex systems. The governance

principles on the other hand, deal with proper implementation of the theoretical

principles on both an individual and organizational scale. The complex and constantly

growing autonomous vehicle system will be analyzed thoroughly in order to make a

recommendation that is effective and practical; as well as ensuring the financial integrity

of Tesla, Inc.


As a recent technological development, autonomous vehicles implement many modern

technology systems. This includes both the software used for navigation, detection, and

more recently, artificial intelligence; as well as the hardware such as the sensors and


cameras. Both the software and hardware will greatly improve the safety of passengers

as opposed to regular vehicles for numerous reasons. At the same time however, new

problems are introduced such as criminal liability, cyber threats, and malfunctions.

According to an ESEM governance principle, “Major shifts in technologies and

technological systems should, to the extent possible, be explored before, rather than

after, implementation of policies and initiatives designed to encourage them” (Allenby,

2013, p.187). A major shift in technology in this case would be the use of artificial

intelligence. Artificial intelligence, especially advanced forms, is still in its initial design

stages. However, it has potential to greatly improve the software used in autonomous

vehicles by creating an intelligent and constantly learning system. This incentivizes

companies to implement artificial intelligence as a part of the autonomous technology

and it is evident that policy surrounding autonomous vehicles have already been

enacted or are in progress. The issue is that the use of artificial intelligence at such a

large and dynamic scale could have potential negative consequences. Artificial

intelligence has not fully been explored at this level, thus there are still many safety and

ethical considerations of this technological system. Since the incentive to design

artificial intelligence based autonomous vehicles already exists, these considerations

cannot be fully accounted for before it becomes commonplace.

According to Danny Shapiro, senior director of automotive at NVIDIA, the technology

used in autonomous vehicles will improve the overall safety of the passengers. The

sensors can detect the environment accurately by building a complete three-

dimensional map, thus minimizing the chance of a potential collision with other vehicles,

pedestrians, cyclists, and obstacles. One way this can be achieved is by using Lidar

technology. Lidar is a type of sensor that detects the location of objects by measuring

the length of time it takes for light to reach the object. It then creates a 3D map made up

of millions of points that the light detected. The software algorithm can use this map to

navigate the environment.


Another way it can improve the passenger’s safety is because the software will follow all

rules and regulations of the road, such as the speed limit, whereas a human driver may

not always do so. A system of inter-communicating autonomous vehicles all obeying

traffic laws have almost no opportunities to make mistakes. Therefore, it is expected

that the rate of accidents in the United States will drop drastically when this becomes a

reality. Conditions that can cause human errors, like tiredness or driving under the

influence, will be completely eliminated as well. This is one of the major benefits of

large-scale autonomous driving systems since nearly every single accident is caused by

something form of human error. Especially recently, since distracted driving is becoming

an increasingly common occurrence due to smart phone use and other electronics.

Solving this problem also allows for the passengers to have uninterrupted free time.

This opens up room for work to be done or to simply rest, and in both cases improving


While this technology will clearly reduce the overall number of accidents, it has several

downsides as well. The two main ones are potential cyber threats as well as system

malfunctions. Cyber threats would be a targeted attack on the system. For instance, a

hacker may use an exploit to gain access to the autonomous driving system. The

possibilities range from controlling minor aspects of the vehicle, such as the lights, or to

full control including navigation. This creates a massive criminal liability issue as the

current laws assume that the person in the driver seat has full control of the vehicle for

legal purposes. Individual states and legislators would have to either amend the current

laws or be “faced with the daunting challenge of creating a new set of regulations that

will satisfy the public need for safety while simultaneously realizing the potential benefits

of autonomous vehicle technology” (Douma, Palodichuk, 2012, p. 1162). More

importantly however, the safety of the passengers and other people on the road is

called into question. Possible large-scale terror attacks can be conducted if a criminal

gains access to an entire network of autonomous vehicles. Although, this is just

speculation at this point in time since there is no precedent for an attack of this

magnitude. Regardless, it is becoming increasingly important for manufacturers to


include a type of failsafe, such as an override button or a way for the software to detect

an intrusion and notify the passengers or proceed with an emergency protocol.

A failsafe will also be important in the case that the system malfunctions or has any sort

of error. As with any complex and advanced technological systems, especially ones that

are in their preliminary stages, they are subject to malfunctions. These malfunctions can

be small such as a minor traffic violation, or large enough that it causes an accident on

a quickly moving freeway. Being put in a life threating situation due to an error or glitch

in the software is one of the reasons that people may be resistant towards autonomous

vehicle technology. The issue is that the only way to minimize system errors and

malfunctions is rigorous testing and quality control of the software and underlying

mechanisms. This will be costly however, as it will take a lot of investment in research

and development and quality assurance. Automotive companies that are developing

autonomous driving technology, such as Tesla, should be prepared to invest heavily in

these two steps particularly.


Autonomous vehicles have many environmental implications to consider. Depending on

certain factors, they can either increase or decrease emissions and thus impact the

carbon footprint. In terms of efficiency however, they are much more efficient than

regular vehicles by improving freeway throughput and removing inefficient human

driving habits. In this case, efficiency refers to the usage of resources, especially natural

gas, over time. Lastly, autonomous vehicles will greatly impact the current infrastructure

of a city as well as planning of upcoming infrastructure to accommodate them better.

According to a theoretical Earths Systems Engineering and Management principle, “The

way problems are stated defines the systems involved” (Allenby, 2013, p.185). Thus,

due to the purposes of this report, it is necessary to frame the environmental problem

from a perspective relevant to Tesla, Inc. The environmental aspects of autonomous

vehicles will be assessed not only though ESEM principles, but also with Tesla’s

corporate social responsibility strategy which takes into account all relevant

stakeholders by priority as given in their business model.


The first environmental impact is vehicle emissions. Many special interest and

environmental groups prioritize emissions because they are a major contributing factor

to greenhouse gasses. The United States Environmental Protection Agency states that

“A typical passenger vehicle emits about 4.7 metric tons of carbon dioxide per year”.

Whether the overall emissions will increase or decrease with the use of autonomous

vehicles is difficult to say without years of data collection and analysis, but there are

several predictions that can be made. The first is that the number of vehicles on the

road and the length of time people spend in their vehicles is very likely to increase with

autonomous driving systems. This is because the opportunity cost of driving will be

lowered, and people that could not drive before such as the elderly, are able to travel in

autonomous vehicles alone. As a result, emissions will increase assuming that these

are still natural gas based vehicles. Tesla is likely to solve this issue by incorporating

their electric vehicle technology with their autonomous vehicle systems, thus satisfying

the community and environmental protection groups. Secondly, due to efficiency

benefits and potential for easy ride-sharing, automation may in fact lower carbon

emissions. This is especially true in the case of long-distance trucks which would greatly

benefit from the increased efficiency of autonomous vehicles.

The efficiency benefits of autonomous vehicles are numerous, with one factor being the

elimination of inefficient human driving habits and errors. The second factor is the

increase in freeway throughput by communicating to other vehicles. Humans cannot

perform as well as an automated driving system. People tend to make avoidable

mistakes, such as missing an exit or making a wrong turn which would inevitably create

more traffic congestion as well as more time spent on the road. A well programmed

automated system cannot make mistakes like this, therefore improving general

efficiency. The second factor is more complicated and relies on various variables but it

can be simulated without the need for field data collection. According to research done

by Abdullah Maarafi “results have shown that incorporation of autonomous vehicles with

regular vehicles can increase the freeway throughput. The increase observed in our

study has reached above 17% of freeway benefits with 60% or higher of autonomous

vehicles penetration rate.” (2015, p. 3). Maarafi draws the conclusion that with more

autonomous vehicles on the road, there will be less congestion on freeways. This


greatly benefits the environment by reducing traffic and carbon emissions. This is

especially true if autonomous vehicles were all on the same network and in continuous

communication, although this technology may be at least a decade away.

To have every vehicle be autonomous and on the same grid, the infrastructure would

have to change too. Roads and highways would need to meet the needs of such a

massive change to the fundamentals of driving. City planning in the future would also be

affected greatly, assuming there will be little to no more regular vehicles in use. With

autonomous vehicles on the road, stoplights may not need to exist in their current form

if all the vehicles are online and in constant communication. This will greatly improve

safety and efficiency by reducing the number of accidents at intersections, and also by

reducing wait time and fuel consumption. Highways may have to be expanded to

include additional lanes as well, since there will be an influx of vehicles on the road due

to reasons previously mentioned. If this proves to be an essential and a common

change, all of the construction work may be detrimental to the environment.

Alternatively, a fully autonomous highway system may not need additional space due to

more efficient traffic flow. Again, it is unclear which will be the case without further data

collection and experimentation. Either way, according to a key theoretical ESEM

principle, intervention by governing bodies in this case should be avoided. This implies

that major rebuilding or construction of infrastructure should not take place until it is

deemed necessary, otherwise it may create an undesirable and unpredictable



Lastly, since autonomous vehicles are a complex ESEM problem, they will undoubtedly

create many socioeconomic impacts. Allenby states, as a theoretical principle, “ESEM

projects and programs are highly scientific and technical in nature—but they also have

powerful economic, political, cultural, ethical, and religious dimensions as well. All of

these facets should be explicitly integrated into ESEM approaches” (2013, p. 185).

Therefore, it is necessary to consider all aspects of autonomous vehicles, and not


simply focus on the technical and engineered parts. This includes, legislation and policy,

economic effect, as well as moral and ethical matters.

The first social impact of course will be the legislation regarding autonomous vehicles.

As with the criminal liability problem, other similar issues will be raised. It will be difficult

to determine who will be at fault in the case of a crash and thus whose insurance

company is liable for the damages, for instance. In general, there will be many specific

regulations dealing with factors such as road types, zones, and environmental

conditions. Each jurisdiction has varying traffic laws in the United States, thus the

autonomous vehicle systems need to account for those when crossing borders to drive

in another state. There needs to be full communication between legislators and

manufacturers before commercially available and full scale autonomous vehicles can

exist. Terms and conditions need to be discussed and definitions need to be made on

what exactly constitutes an autonomous vehicle. Most importantly, the legislators should

clearly list every requirement that an autonomous must fulfill before it is able to be sold

to the general public, even basic ones such as yielding and parking legally. This

“creates a clear and transparent expectation among both legislators and manufacturers

about the challenges AVs must surmount in order to become commercially salable”1.

This explains that the main motivation for communication with legislators is to create a

clear and open dialogue with no misunderstandings. Of course, other administrative

details have to be established as well. One such detail is the licensing for a private

individual to own and operate such vehicles on public roads.

1University of Washington’s Technology Law and Policy Clinic

The economic impact of commercially available autonomous vehicles will be severe and

extend across many different industries. The main one being the automotive industry

since “the number of vehicles purchased each year may fall, due to vehicle-sharing

within families/across household members or through shared fleets, but rising travel

distances and a shift away from air travel may lead to greater vehicle-miles traveled

(VMT) and ultimately higher vehicle sales (due to faster fleet turnover from heavy daily

use).” (Kockelman, Clements, 2017, p. 1). Both these scenarios are possible and will


cause the market to shift in either direction. This is something that Tesla needs to

account for in order to stay profitable. Another important factor is the value of hardware

relative to software. “[Connected and fully Autonomous Vehicles] will soon be central to

the automotive industry, with software making up a greater percent of vehicle value than

it had previously and hardware’s percentage value falling” (Kockelman, Clements, 2017,

p. 1). It is becoming evident that a shift towards the software that powers autonomous

vehicles will be necessary. The car itself and the hardware components such as the

camera and sensors, will not be as important as having a quality tested and effective

software system. Once such a system is established and improving it begins

approaching diminishing returns, then focus can shift back to design and manufacture of

the hardware involved in autonomous vehicles.

Before full commercialization, autonomous vehicles will likely be established with the

most simple and noticeable case; heavy truck driving and distribution. This will cause

job loss across that entire industry but in turn reduce costs for shipping and towing

companies greatly by eliminating inefficient human driving, like the need to take breaks.

Another industry that is likely to adopt autonomous vehicles quickly is the ride-sharing

economy. Services that offer people rides to get to their desired location will greatly

benefit from automated vehicles as it will create no need to pay a human driver and of

course all of the aforementioned efficiency savings. Public methods of transportation,

such as taxis and busses will be forced to adopt autonomous vehicles as well or they

will simply become redundant. Both these public and private methods of transportation

account for a significant portion of American employment. Automating these jobs may

indirectly create an employment issue which should also be considered as per ESEM

practice. Although an autonomous vehicle industry will certainly create more jobs, it will

mostly be skilled labor positions like computer programmer or designer. Thus, all of

these displaced low skill laborers will have difficulty finding jobs that they can work,

adding to the unemployment issue in the United States.

Finally, autonomous vehicles raise a multitude of ethical and moral considerations. The

main one being deciding which system of morality will govern the software’s decision

making. Morals will differ across distinct cultures, countries, and time periods. It is


problematic to choose one underlying system of morality that encompasses all views

because a perfect system may not exist. For instance, in an unavoidable collision,

would the autonomous driving software choose to prioritize the safety of the passengers

over a pedestrian, the pedestrian over the passengers, or would it just leave it up to

chance. This type of question has no obvious answer since it will of course depend on

the ethical values of the person answering it. Another difficulty arises is in programming

the vehicle to break the law when necessary. A scenario may arise that in order to avoid

a collision with a living thing, it must make an illegal turn but the software is

programmed to never break the law. Exceptions would have to be created and it would

get increasingly complicated to consider all possible scenarios in which the vehicle may

break the law ethically. It once again creates additional questions of how should the life

of living things be prioritized. The Trolley Problem thought experiment is an example

that deals with this, and the general consensus is to take the utilitarian approach but this

is once again not a universally accepted solution. Another major difficulty will be in

regard to implementation of a thorough moral and contingency system as code in the



In conclusion, autonomous vehicles have a considerable number of technological,

environmental, and social implications. These vehicles introduce both emerging software and

hardware capabilities that still need to be fully explored. It is well documented that there will be

an overall reduction in accidents due to these advanced technology systems. They do however,

impose a security threat that can at least be partially solved by including failsafe mechanisms.


Environmentally, emissions may or may not be reduced depending on several factors. If Tesla,

Inc. implements their electric vehicle technology, then carbon emissions from vehicles due to

natural gasses will be entirely eliminated. Also, autonomous vehicles have shown to increase

efficiency in two main ways. The first being freeway throughput, and the second being the

removal of inefficient human driving habits. These improvements come by nature of an

autonomous driving system.

Lastly, the impacts that autonomous vehicles have on the social sphere are reasonably

predictable in the case of economy, but uncertain in terms of ethics and morality. Creating a

governing ethical decision-making system that a majority of people will accept will prove to be a

difficult task. As for legislation and policy, law makers would have to adapt to autonomous

vehicle technology quickly. The industry is growing fast and appropriate rules and regulations

need to be in place before mass commercialization.

Overall, the goals and objectives of this report were met. The technological, environmental, and

social impacts of autonomous vehicles were assessed through the lens of ESEM principles.

Relevant recommendations were then made to Tesla, Inc. that take into account these three

major aspects of the autonomous driving system while adhering to ESEM principles. Tesla will

need to focus on creating an autonomous driving system that is safe, secure, ethical, and

effective while also maintaining their corporate social responsibility and profitability as a publicly

traded firm.


Allenby, B. …