Proceedings

Term Papers prepared for

810:161, Artificial Intelligence, Spring 2009

University of Northern Iowa

(UNIAI-09)

Artificial Intelligence & Autonomous Vehicles

• Autonomy in Exploring the Unknown : A Survey
  Brittany Elliott and Benjamin Fain   (full paper)
  

  This paper presents various information about the recent uses and developments of artificial intelligence in the exploration of unknown environments, specifically underwater and space exploration. The use of autonomy in vehicles with regards to exploring unknown areas is very important because oftentimes direct human influence cannot easily be maintained with these vehicles. These vehicles must be able perform tasks dynamically like planning and re-planning their own activities based on external factors, recognizing and exploring their environments without the aid of humans, and scientifically analyzing any interesting data or specimens they may encounter. Autonomous Underwater Vehicles use methods such as SLAM (Simulated Localization and mapping) to observe and map its surroundings while some Martian rovers use systems such as OASIS(Onboard Autonomous Science Investigation System), CASPER(Continuous Activity Scheduling, Planning Execution and Replanning), and ASE(Autonomous Sciencecraft Experiment) to guide their missions. Finally, we will provide some information on the future hopes and expectations of autonomy in this field.

   

• Flight Control via AI : A Problem Analysis
  Cody Johnson, Allyn Bauer, Leandro Avila  (full paper)
  

  With the amount of air traffic growing ever day, it has become clear that we need a solution to help the people that manage the skies. One solution to this uses multiple agents that are trained using machine learning to help. This is a partially complex problem, there needs to be many safety checks and a black box solution would not be acceptable. The algorithms that these agents run on must also be fast and efficient because air traffic control is a high speed world where decisions must be made in seconds, there is no time to wait if the agent has to spend a great deal of time calculating the solution. The number of flights in the nation is increasing every day. This paper will discuss the problems involved in creating an agent to act as an air traffic controller. First by talking about what is involved in air traffic control and in what areas the agents can help. Second, the techniques that could be used to implement the agents for this role will be addressed.

   

• Techniques Used in Autonomous Vehicle Systems : A Survey
  J.C. Last, and Robbie Main   (full paper)

  This paper focuses on a variety of techniques implemented in urban autonomous vehicles. Four major areas of autonomous vehicle research are categorized into the follow sections: recognizing the road and staying on it, recognizing objects and dealing with them, handling different types of terrain, and awareness and goals. Each section is broken down further, answering common, specific questions associated with each area. Discussion will include examples of and how different techniques, methods, and models work. These various techniques are used in vehicle autonomy given different situations.

   

• Analysis of Unmanned Combat Air Vehicles
  Curtis Seitz and Brian Kroeze  (full paper)

  There may be a time in the near future when the majority of our airborne campaigns are completed through the use of Autonomous vehicles. Unmanned combat air vehicles (UCAV) could be the main instrument in implementing these military and reconnaissance missions. Research in the field of UCAV can be broken up into many different specific domains that come together to complete the UCAV. Throughout the course of this paper we will be analyzing the different components that make up a UCAV system. We will be analyzing the different systems that need to be perfected to have a useful UCAV.

   

   

Game Playing

• Chess Playing Agents : A Problem Analysis
  Francisco Mota and Alex Rottinghaus   (full paper)

  In this article we discuss chess as a problem to be solved by artificial intelligence. We analyze chess and describe why it is a particularly popular AI problem. This article also explores the history of chess playing agents and discusses their advancement over time. We also highlight the competition that erupted between chess players and computer engineers. We also discuss different algorithms and programming techniques for playing chess. We then offer an opinion of the combination of techniques that are most often used and seem to have the greatest success at making winning chess agents.

   

• Evolution of Artificial Intelligence in Video Games : A Survey
  Ken Mott   (full paper)

  Artificial Intelligence in video games has grown greatly over the recent years. From the basic Tracking AI of Atari’s Pong to the table of locations used in Sony’s Killzone for the Playstation 2. The style of AI used varies greatly between genres of games. There is a large difference between how the AI in a First Person Shooter acts as opposed to the AI in a Real Time Strategy game. There is also advancements in video game AI to make characters more life-like, called A-Life. The future of AI in video games is bright. More time is being spent in game development working on making good AI than ever before. This paper will go into much more detail on all of these topics.

   

• Artificial Intelligence in Modern Video Games : A Survey
  Sam Palacek and Ryan Watermiller   (full paper)

  Video games continue to evolve at an incredible rate, and with this comes advances in technology and the artificial intelligence behind them. Artificial intelligence is highly involved in many aspects of game design, and along with the graphics, is responsible for the believability of any game. In this paper we will discuss many different aspects of AI in the video games of today. Some things we will discuss include: skill levels of computer players, gaming environments, and realness of the computer agents. We will also discuss pathfinding methods that are used in many games, and how they can be improved with new algorithms and heuristics.

   

• Poker Computers and Chess Computers : A Tutorial
  Lukus Reindl and Patrick Paulsen   (full paper)

  The following paper describes the fundamental elements of chess and poker computers. It goes over how chess computers have evolved into what they are today, taking on the best human chess players in the world. Poker computers today are growing more and more popular among researchers. While chess computers have had a relatively large amount of research poker is becoming the new frontier for artificially intelligent computer games. Included are algorithms, such as search trees, used in the design of chess computers. Along with this are design aspects of poker computers, such as generating data models of opponents’ behavior. Lastly a brief description of the ultimate difference in these two fields of research.

Robotics and Multi-Agent Systems

• Human-Robot Interactions : A Survey
  Mohammed Al Baharnah and Samantha L. Fahrmann   (full paper)

  As technology develops, scientists create new ways to make robots useful in different environments. Robots have begun to function with humans and interact with them on a limited basis. Robots use operating systems that contain software with design patterns specific to each application. Robots contain sensors that give input and mechanisms that give different forms of output. In human-robot interaction, the interaction is a compound of different forms of communication. The purpose of this paper is to focus on research covering different aspects of human-robot interaction. There is a great deal of discussion on the future of human-robot interaction based on its application in everyday life.

   

• Heuristic Path Planning Algorithms : A Survey
  Eddie Maldonado and Josh Mains  (full paper)

  This paper consists of a survey of heuristic path planning algorithms. The discussion will be focused on four different topics. These topics will be path planning algorithms, incremental re-planning algorithms, anytime algorithms, and finally re-planning anytime algorithm. For path planning algorithms the paper looks at two different algorithms. These algorithms are A* and Greedy Localization. For incremental re-planning algorithms two separate algorithms will be investigated. These two separate algorithms are Bug Modified 1 and D* Lite. For anytime algorithms there are two algorithms that will be investigated. These algorithms are Anytime A* and Learning Real-Time A*. Lastly, for re-planning anytime algorithm the paper looks at an algorithm that is a compilation of multiple algorithms, and this algorithm is known as Anytime Dynamic A*.

   

• Robotic Vision Systems : A Survey
  Keller J.C. McBride and Trent Johnson   (full paper)

  This paper is a survey of robotic vision systems. We will cover a wide array of angles of robotic vision systems beginning with an introduction defining the concepts behind these vision systems and a history of these systems in general. We will move on to cover some of the techniques that are used and a few specific examples that show some of the systems that robotic vision has been integrated into.

Innovative Applications of Artificial Intelligence

• Signal Processing : Modern Applications in Genetics and Music : A Survey
  Dustin Collins and Scott Smart   (full paper)

  In this paper we discuss the significance of artificial intelligence in modern signal processing, from search applications to projects in genetics and music. Implementation varies by the type of signal involved, but technologies concerned with reading signals share common foundations. The difference in analyzing digital and analog signal is discussed, as well as string and genome processing applications. Music analysis practices are touched on, as well as sample projects arising from the technology. Finally, learning is discussed with respect to music creation in a real-time environment.

   

• A Survey of Intelligent Solutions in Network Management and System Administration
  J. Schweitzer, A. Hamad, and M. Karl   (full paper)

  This paper is a survey of robotic vision systems. We will cover a wide array of angles of robotic vision systems beginning with an introduction defining the concepts behind these vision systems and a history of these systems in general. We will move on to cover some of the techniques that are used and a few specific examples that show some of the systems that robotic vision has been integrated into.This paper investigates some techniques and implementations of integrating artificial intelligence into system administration and network management. We divide system and network administration into three main categories: daily tasks, system and network health, and system and network security. We examine two implementations of artificial intelligence used for daily system administration tasks, one developed by J. Hamlin and W. Potter from the Artificial Intelligence Center at the University of Georgia and another developed by F. Koch and C. Westphall. We then look at a network-probing technique by Brodie, Rish, and Ma of IBM and some prediction techniques by Vilalta, Apte, Hellerstein, Ma, & Weiss, also of IBM, that help in monitoring the health of systems and networks. Along with this we briefly investigate a tool called ABLE that allows for easy creation of complex, multiagent intelligent systems. Finally, we examine a few inplementations of artificial intelligence in system and network security. We look at a host-based IDS called Tripwire developed by Purdue University and SnortAI, a plug-in for a network-based IDS called Snort.