# Artificial Intelligence Future Directions Computer Science Essay

Published:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

The most famous historical fictional car in Knight Rider TV series which can drive by itself by making decisions, learn, communicate and interact with humans is now becoming a reality today due to the emergence of technology or future directions of Artificial Intelligence. This literature survey is about one of the interesting topic of Artificial Intelligence that is Unmanned driving ground vehicles, future direction of Artificial Intelligence. Chapter 1 gives an introduction to the topic, brief history and current status of the emerging area. Chapter 2 discuss about related areas of Artificial Intelligence used and there theoretical background. Chapter 3 draws attention towards currently available applications of those AI concepts, mainly about driver-less vehicles and their strengths, technologies used by them. Finally Chapter 4 the conclusion conclude the survey by giving current position of driver-less vehicles and shiny future of driver-less vehicles.

\end{abstract}

\newpage

% Abstract ends

% Acknowledgement begins

\begin{center}

\section*{Acknowledgements}

\end{center}

\paragraph{}

This thesis stands here today due to remarkable assistance given by many people. First I would like to thanks Dr. Lalith Premeratne for the grateful support given by him during the period of the literature survey. Giving me the chance to decide on any project which I wish to do. Also gave me the assistance how to structure my final report. And I would like to thank Mr.Ashoke Ekanayake and Mr.Asankha C Perera for the support given for this thesis to make this success. Also my parents support and encourage me also through out this course. Also I would like to thank Samanmalie, Duminda, Samudini for the help given in this course. I am grateful to all every one who helped me throughout this survey.

\newpage

% Acknowledgement ends

\tableofcontents

\listoffigures

\listoftables

\newpage

% Abbreviations ends

% Chapter 1 begins

\chapter{Introduction}

\section{What is ?}

\paragraph{}

What is a driver-less car? A driver-less car is an autonomous vehicle that can drive itself from one point to another without using any assistance from a driver, with the use of an autopilot system. Autonomous vehicles have the potential to transform the transportation industry while virtually eliminating accidents and cleaning up the environment. Main impetuses behind the driver-less car is safety.

\paragraph{}

Driver-less car has to perform functions followed by skilled human driver. It has to sense, decide and act. So the software inside should be capable of acquire sensor data, construct internal models of the environment and make driving decisions.

\paragraph{}

In order autonomous vehicle to drive it need following needs

\begin{enumerate}

\item Sensors - to understand its immediate environment

\item Navigation - know what its current position and destination

\item Motion Planning - find out its way in traffic or obstacles

\item Actuation - operate the mechanics of the autonomous vehicle

\end{enumerate}

\paragraph{}

Currently $2\frac{1}{2}$ of these problems are already solved. Navigation and Actuation related problems completely, problems of Sensors partially, but improving rapidly. The main unsolved part is the problem of Motion Planning.

\section{Motivation}

\paragraph{}

Its one of the dream of man kind to build cars that drive by themselves. Also annually several thousand of valuable lives lost 41,000 only in United States alone due to traffic accidents all over the world. What could have prevented the accident? And the obvious answer is you by paying attention while driving. But that answer is not so simple as it seems. Driver error is the most common cause of traffic accidents, because of cellular phones, vehicle entertainment systems, complicated road systems, more traffic or tension in their mind result in loosing concentration. Lost concentration for two seconds may loose your whole life. So if drivers are not going to concentrate on the road, who is? Nothing else thats your car. Due to the development of technology car makers are focusing on systems that can make cars safe, even if drivers are not. Driver less vehicles for the rescue.

\pagebreak

\paragraph{}

Also autonomous vehicle have the potential to transform the transportation industry while eliminating accidents. More and more private cars generate more and more traffic, also which are not in use and parked 90\% of the time. By using public self driving transportation will reduce traffic meanwhile reducing accidents. This would also allow getting appropriate vehicle for the particular need.

## \\

\begin{figure}[h]

\begin{center}

\includegraphics[scale=.7]{self-drive-taxi.jpg}

\caption{public self driving taxi}

\label{picture}

\end{center}

\end{figure}

\section{History}

\paragraph{}

Following is the brief history of autonomous driver-less vehicles.

\begin{table}[h]

\caption{Highlights of Autonomous Vehicle History\cite{4,2,6}}

\begin{center}

\begin{tabular}{|>{\bfseries}l| p{12.35cm}|}

\toprule

\hline Year & \textbf{Description}\\

\hline 1977 & A vehicle of Japan's Mechanical Engineering Lab follows roads for up to 50 m at up to 30 km/h.\\\midrule

\hline 1980s & Prof. Ernst Dickmanns and his group at Bundeswehr University of Munich, Germany build the world's first real robot cars, using saccadic vision, probabilistic approaches such as Kalman filters, and parallel computers. A vision-guided Mercedes-Benz robot van, designed by Ernst Dickmanns and his team achieved 96 km/h on streets without traffic.\\\midrule

\hline 1987-1995 & European Commission began funding the 800 million Euro EUREKA Prometheus Project on autonomous vehicles. \\\midrule

\hline

\end{tabular}

\end{center}

\end{table}

\pagebreak

\begin{table}[h]

\begin{center}

\begin{tabular}{|>{\bfseries}l| p{12.35cm}|}

\toprule

\hline Year & \textbf{Description}\\

\hline 1980s & DARPA-funded Autonomous Land Vehicle (ALV) in the United States achieved the first road-following demonstration that used laser radar (Environmental Research Institute of Michigan), computer vision (Carnegie Mellon University and SRI), and autonomous robotic control (Carnegie Mellon and Martin Marietta) to control a driver-less vehicle up to 30km/h.\\\midrule

\hline 1987 & HRL Laboratories (formerly Hughes Research Labs) demonstrated the first off-road map and sensor-based autonomous navigation on the ALV. Vehicle traveled over 600m at 3 km/h on complex terrain with steep slopes, ravines, large rocks, and vegetation.\\\midrule

\hline 1994 & The twin robot vehicles VaMP and Vita-2 of Daimler-Benz and Ernst Dickmanns of UniBwM drove more than 1000 kilometers on a Paris three-lane highway in standard heavy traffic at speeds up to 130 km/h, albeit semi-autonomously with human interventions. They demonstrated autonomous driving in free lanes, convoy driving, and lane changes left and right with autonomous passing of other cars.\\\midrule

\hline 1995 & Dickmanns re-engineered autonomous S-Class Mercedes-Benz took a 1600 km trip from Munich in Bavaria to Copenhagen in Denmark and back, using saccadic computer vision and transputers to react in real time. It achieved speeds exceeding 175 km/h on the German Autobahn, with a mean time between human interventions of 9 km, or 95\% autonomous driving. Again it drove in traffic, executing maneuvers to pass other cars. Despite being a research system without emphasis on long distance reliability, it drove up to 158 km without human intervention.\\\midrule

\hline 1995 & Carnegie Mellon University Navlab project achieved 98.2\% autonomous driving on a 5000 km (3000-mile) "No hands across America" trip. This car, was semiautonomous by nature: it used neural networks to control the steering wheel, but throttle and brakes were human-controlled.\\\midrule

\hline 1996-2001 & Prof. Alberto Broggi of the University of Parma organized the ARGO Project, which worked on enabling a modified Lancia Thema to follow the normal (painted) lane marks in an unmodified highway. The culmination of the project was a journey of 2,000 km over six days on the motorways of northern Italy dubbed MilleMiglia in Automatico, with an average speed of 90 km/h. 94\% of the time the car was in fully automatic mode, with the longest automatic stretch being 54 km. The vehicle had only two black-and-white low-cost video cameras on board, and used stereoscopic vision algorithms to understand its environment, as opposed to the "laser, radar - whatever you need" approach taken by other efforts in the field. \\\midrule

\hline 2002 & the DARPA Grand challenge competitions were announced. DARPA is a prize competition for driverless vehicles, funded by the Defense Advanced Research Projects Agency, the most prominent research organization of the United States Department of Defense.\\\midrule

\hline 2004-2005 & DARPA competitions allowed international teams to compete in fully autonomous vehicle races over rough unpaved terrain and in a non-populated suburban setting.\\\midrule

\hline 2007 & DARPA challenge, involved autonomous cars driving in an urban setting.\\\bottomrule

\hline

\end{tabular}

\end{center}

\end{table}

\pagebreak

\section{Current Status}

\paragraph{}

Following are the current approaches in autonomous vehicles area

\begin{itemize}

\item Fully autonomous driving vehicles

\item Prebuilt infrastructure for automated vehicles

\item Driver assistance that remove some requirements from human driver

\end{itemize}

\paragraph{}

An important concept that cuts across several of the efforts is vehicle platoons. In order to better utilize road space, vehicles are assembled into ad-hoc train-like "platoons", where either human or automatic driver of the first vehicle makes all decisions for the entire platoon. All other vehicles simply follow the lead of the first vehicle.

\paragraph{}

Driver less vehicles fall into fully autonomous driving vehicle category. In recent times DARPA (Defense Advanced Research Project Agency) challenge is the main competition which assist research teams in many different universities to come with better solutions in driver less vehicles while competing for \$2 million grand prize-money. \subsection{The DARPA Urban Challenge} \paragraph{} DARPA is a major competition(large part a software competition) for driver-less vehicles, funded by the Defense Advanced Research Projects Agency, the most prominent research organization of the United States Department of Defense. Some of this future driver-less car technologies may be closer than you think. The DARPA Urban Challenge pits teams against each other to create cars that can negotiate traffic autonomously. The goal of the program is not just to reduce traffic accidents and congestion, however: it is to produce driver-less vehicles for combat, keeping soldiers far from the front lines. And the winner will receive \$2 million. Following is the brief statistical data about 2004, 2005 and 2007 DARPA challenges.

\subsubsection{DARPA Grand Challenge 2004}

\paragraph{}

The very first competition was held on March 13, 2004 in the Mojave Desert region of the United States, along a 150-mile (240 km) route. None of the robot vehicles finished the route. Carnegie Mellon University's vehicle traveled the farthest distance, completing 11.78 km (7.36 miles) of the course. The red team won that year. But the grand prize remained unclaimed.\cite{7}

\begin{figure}[h]

\begin{center}

\includegraphics[scale=.5]{red-team1.jpg}

\caption{\textit{Sandstorm} winner of 2004 DARPA created by Red Team from Carnegie Mellon University \cite{7}}

\label{picture}

\end{center}

\end{figure}

\pagebreak

\subsubsection{DARPA Grand Challenge 2005}

\paragraph{}

The second competition was a 212 km (132 mile) off-road course that was held on October 8, 2005. All but one of the 23 finalists in the 2005 race surpassed the distance completed by the best vehicle in the 2004 race. Five vehicles successfully completed the race. \cite{8}

\begin{table}[h]

\caption{2005 DARPA Grand Challenge Summary \cite{8}}

\begin{center}

\begin{tabular}{|>{\bfseries}l| p{2.5cm}|p{5cm}|p{2cm}|p{2.5cm}|}

\toprule

\hline Vehicle & \textbf{Team Name} & \textbf{Team Home} & \textbf{Time Taken(h:m)} & \textbf{Result}\\

\hline Stanley & Stanford Racing Team & Stanford University, Palo Alto, California & 6:54 & First place\\\midrule

\hline Sandstorm & Red Team & Carnegie Mellon University, Pittsburgh, Pennsylvania & 7:05 & Second place\\\midrule

\hline H1ghlander & Red Team & Carnegie Mellon University & 7:14 & Third place\\\midrule

\hline Kat-5 & Team Gray & The Gray Insurance Company, Metairie, Louisiana & 7:30 & Fourth place\\\midrule

\hline TerraMax & Winner of 2004 DARPA Team TerraMax & Oshkosh Truck Corporation, Oshkosh, Wisconsin & 12:51 & Fifth place Over 10 hour limit\\\midrule

\hline

\end{tabular}

\end{center}

\end{table}

\subsubsection{DARPA Grand Challenge 2007}

\paragraph{}

The third competition commonly known as the DARPA Urban Challenge held in November 3, 2007 at the site of the now-closed George Air Force Base in Victorville, California. The course involved a 96 km (60-mile) urban area, to be completed in less than 6 hours. Should have to obeying all traffic regulations while negotiating with other traffic and obstacles and merging into traffic. Initially 53 teams qualified but only six teams successfully completed the course. \cite{9}

\begin{table}[h]

\caption{2007 DARPA Grand Challenge Summary \cite{9}}

\begin{center}

\begin{tabular}{|>{\bfseries}l| p{2cm}|p{5cm}|p{2.2cm}|p{3cm}|}

\toprule

\hline Vehicle & \textbf{Team Name} & \textbf{Team Home} & \textbf{Time Taken(h:m:s)} & \textbf{Result}\\

\hline Boss & Tartan Racing & Carnegie Mellon University, Pittsburgh, Pennsylvania & 4:10:20 & 1st Place; average speed 14 mph (22.53 km/h)

\\\midrule

\hline Junior & Stanford Racing & Stanford University, Palo Alto, California & 4:29:28 & 2nd Place; average speed 13.7 mph (22.05 km/h)\\\midrule

\hline Odin & VictorTango & Virginia Tech, Blacksburg, Virginia & 4:36:38 & 3rd Place; average speed less than 13 mph (20.92 km/h)

\\\midrule

\hline Talos & MIT & MIT, Cambridge, Massachusetts & Approx. 6 hours & Fourth place\\\midrule

\hline Little Ben & The Ben Franklin Racing Team & University of Pennsylvania, Lehigh University,Philadelphia, Pennsylvania & No official time & One of 6 teams to finish\\\midrule

\hline

\end{tabular}

\end{center}

\end{table}

\pagebreak

\begin{figure}[h]

\begin{center}

\includegraphics[scale=1.55]{stanely7.jpg}

\caption{\textit{Stanely} winner of 2005 DARPA created by Stanford Racing Team from Stanford University \cite{8}}

\label{picture}

\end{center}

\end{figure}

\begin{figure}[h]

\begin{center}

\includegraphics[scale=.7]{Boss.jpg}

\caption{\textit{Boss} winner of 2007 DARPA created by Tartan Racing Team from Carnegie Mellon University \cite{9}}

\label{picture}

\end{center}

\end{figure}

\section{Future Plans}

\paragraph{}

The remainder of the literature survey will be focused on identify the areas of Artificial Intelligence used in unmanned driving vehicles or driver-less vehicles. Next review existing real world applications rather driver-less vehicles and their capabilities. Finally the conclusion at the end.

% Chapter 1 ends

\pagebreak

% Chapter 2 begins

\chapter{Areas of Artificial Intelligence}

\section{Areas at Glance}

\paragraph{}

Autonomous driver less vehicles is a fast developing area in Artificial Intelligence and many theories and technologies of Artificial Intelligence used in order to navigate through obstacles to the final destination. Following are the some of the areas of Artificial Intelligence in autonomous driver less vehicles

\begin{itemize}

\item Practical Search Techniques in Path Planning for Autonomous Driving

\item Fuzzy logic techniques for autonomous vehicle navigation

\item Fuzzy Logic in Automated Vehicle Control

\item Using Spatial Label Propagation to Learn Long-Range Traversability

\item Autonomous Path Tracking Using Recorded Orientation and Steering Commands

\item Autonomous Driving approaches Downtown-Computer Vision

\item Online Speed Adaptation using Supervised Learning for High-Speed, Off-Road Autonomous Driving

\item A Self-Supervised Terrain Roughness Estimator for Off-Road Autonomous Driving

\end{itemize}

\section{Path Planning for Autonomous Driving}

\paragraph{}

Practical path-planning algorithm generates smooth paths for an autonomous driving vehicle operates in an

unknown environment, where obstacles are detected by the autonomous vehicle's sensors by through out its path. First variant of the well known A* search algorithm is used, applied to the 3D kinematic state space of the vehicle, with a modified state-update rule that captures the continuous state of the vehicle in the discrete nodes of A* , thus guaranteeing kinematic feasibility of the path. Secondly improves the quality of the solution via numeric non-linear optimization, leading to a local (and frequently global) optimum. Stanford Racing Team's autonomous vehicle, Junior, uses these path-planning algorithms. Junior demonstrated flawless performance in complex general path-planning tasks such as navigating parking lots and executing U-turns on blocked roads, with typical full cycle re planing times of 50-300ms.

\pagebreak

\begin{figure}[h]

\begin{center}

\includegraphics[scale=.6]{junior1.jpg}

\caption{\textit{Junior} using path planning algorithm to traverse}

\label{picture}

\end{center}

\end{figure}

% Chapter 2 ends

% Chapter 3 begins

\chapter{Applications}

\section{Stanley}

\begin{figure}[h]

\centering

\subfloat[]{\label{fig:gull}\includegraphics[scale=.85]{stanely7.jpg}}

\subfloat[]{\label{fig:tiger}\includegraphics[scale=.2]{stanely3.jpg}}

\subfloat[]{\label{fig:mouse}\includegraphics[scale=.21]{stanely2.jpg}}

\caption{Stanley the first autonomous vehicle to win DARPA grand prize}

\label{fig:Stanley}

\end{figure}

\begin{wrapfigure}{l}{}

\begin{center}

\includegraphics[scale=.4]{stanely-radar.jpg}

\end{center}

\caption{Sensor system in Stanley}

\end{wrapfigure}

\paragraph{}

Stanley is a diesel-powered Volkswagen Touareg R5 with four wheel drive, variable height air suspension, and automatic, electronic locking differentials. To protect the vehicle from environmental impact it has been modified with skid plates and a reinforced front bumper. Stanley is actuated via a drive-by-wire system developed by Volkswagen of America's Electronic Research Lab. It has electronic steering control also individual wheel speeds and steering angle, are sensed automatically and communicated to the computer system through a CAN(Controller Area Network) bus interface.

\paragraph{}

Stanley's most of the sensors are mounted on a custom made roof rack. Five SICK laser range finders placed pointing forward towards the driving direction of the vehicle. Color camera for long-range road perception, two forward-pointed antennae of a RADAR system , primary Global Positioning System (GPS) antenna, two antennae for the GPS compass, and various tracking antennae required by the race organizer. A 6-DOF inertial measurement unit is mounted in the trunk. All sensors acquire environment data at rates between 10 and 100 Hertz.

\pagebreak

\begin{figure}[h]

\begin{center}

\includegraphics[scale=.25]{stanely4.jpg}

\caption{Software flowchart which shows the six functional functional groups of the \textit{Stanley}}

\label{picture}

\end{center}

\end{figure}

\paragraph{}

Stanley's computer system is located in the back of the vehicle's trunk in a shock-mounted rack and it carries an array of six 1.6 GHz Pentium M blade computers, a Gigabit Ethernet switch, and devices that connect physical sensors to the system. Also it has a custom-made power system with backup batteries and a switch box that enables the robot to power-cycle individual system components through software. All computers in Stanley operated using Linux operating system. \cite{10}

\paragraph{}

Stanley follows supervised learning as it takes data from Human Drivers and avoids obstacles. Stanley's overall software system consist of about 30 modules executed in parallel. The system is broken down into six layers which correspond to fuctions such as sensor interface, perception, control, vehicle interface, user interface, and global services.

\paragraph{}

Before the DARPA Grand Challenge, Stanley drove more than 1,000 miles through desert terrain in the southwestern U.S. Longest single stretch of autonomous motion was 200 miles along a large cyclic dirt track, on the way it encountered approximately 90 frontal obstacles at speeds of up to 35 mph. Before the race, Stanley drove 418 miles errors free that would have required human intervention. Stanley fished the DARPA Grand Challenge in 6 hours 53 minutes and 58 seconds, ahead of any other autonomous vehicles in the race. The maximum speed was 38.0 mph and the average speed was 19.1 mph, although early in the race the robot averaged 24.8 mph. The robot was paused twice by the

race organizers, for a total of 9 minutes and 20 seconds pause time. And finally ended as the first by passing top-seeded vehicle in the race. \cite{10}

%\pagebreak

## %

%\section{Boss}

\pagebreak

\section{Shelly To Pikes Peak}

\begin{figure}[h]

\centering

\subfloat[]{\label{fig:gull}\includegraphics[scale=.365]{audi1.jpg}}

\subfloat[]{\label{fig:tiger}\includegraphics[scale=.39]{audi3.jpg}}

\subfloat[]{\label{fig:mouse}\includegraphics[scale=.33]{audi4.jpg}}

\caption{Shelly the high performance Audi TTS}

\label{fig:audi tts}

\end{figure}

\paragraph{}

Shelly is an Audi TTS which is the latest creation of Stanford University researchers who are developing technology that could help make driving safer and one day allow ordinary vehicles to drive on their own. It can traverse rough terrain, accelerate quickly and negotiate sharp turns like other high performance sports cars but the difference is there is no driver.

\paragraph{}

Volkswagen, which makes Audi vehicles and is working with Stanford on the Shelley project, has set a goal of creating fully autonomous vehicles by 2028, said Marcial Hernandez, a senior engineer at Volkswagen's electronics research lab in Palo Alto.

Shelly is equipped with Global Positioning System (GPS) receivers and can be programmed to follow any route using a digital map. The research team has developed computer algorithms that let the car make real-time adjustments to terrain and calculate how fast it can go without spinning out of control.

\paragraph{}

It is named Shelly after Michelle Mouton, the first women to win Pikes Peak. It will face its biggest test at Colorado's Pikes Peak, home of the world-famous International Hill Climb that has bedeviled professional drivers with its steep grades and treacherous switchbacks since 1916. At Pikes Peak, Shelley will climb 4,721 feet (1,439 meters) up the 14,110-foot (4,300-meter) mountain on paved and gravel roads as it covers the 12.4-mile (20-kilometer) race course and its 156 turns at high speeds. The feat has never been tried by an autonomous vehicle.

\paragraph{}

Shelley has already reached speeds of 130 mph (209 kph) at Bonneville Salt Flats in Utah. In May, Gerdes' team plans to take Shelley to El Mirage Dry Lake in Southern California and attempt to break the world record for fastest autonomous land vehicle.

% Chapter 3 ends

% Chapter 4 begins

\chapter{Conclusion}

\paragraph{}

Normally people like to drive their cars, driving can be relaxing and freeing also there are still lot of bugs to work out of driver less car technology before people will be ready to trust it because it is an critical application, existence of bug may have to pay with human lives.

\paragraph{}

Driver less technology will always make the call to protect the car and people in it. But a human driver might make a call that does damage to his car and save the lives of others. For example in a critical situation driver less car may decide to steer through people because it does not recognize people it just see a path of least resistance and steer the car towards it. Those kind of extreme issues should be resolved before get into a car and tell it where to go.

\paragraph{}

Before putting driver less vehicles in the market there are some issues to be addressed such as

\begin{itemize}

\item Getting people to trust the car

\item Getting legislators to permit the car into public roads

\item Untangling the legal issues of liability for any mishaps with no person in charge

\item Getting people to give up their freedom to drive wherever they want

\end{itemize}

\paragraph{}

Some driver less technologies such as such as automatic parallel parking and adaptive cruise control, which lets a car handle stop-and-go traffic on its own are already in the market.

General Motors and Carnegie Mellon University researchers are working together to develop driver-less cars and hope to bring one to the market by 2018. Finally the driver less car technology point to improved safety, decreased traffic and decreased pollution because of the reduced traffic.

% Chapter 4 ends

\bibliographystyle{plain}

\bibliography{2007cs143.bib}

\nocite{*}

\end{document}