Seminar: Technical Topics

Course Content:

Reading, review, and discussion of the current literature of computer science and industry professional and technical journals; oral presentations.
Papers for the Course:

AI: Overcoming Technical Challenges

Automotive Networks-Powerpoint Presentation
Hosted by www.Geocities.ws



AI:Overcoming Technical Challenges


This article discusses an Artificial Intelligence meeting held to seek deployed applications and emerging applications in Artificial Intelligence. These applications are submitted so that application developers can receive guidance for their applications and researchers can better understand technical challenges provided by real-world problems.

At IAAI-2001, six application papers were selected. These authors were asked to expand on their papers to further describe Artificial Intelligence challenges. These articles �provide an excellent snapshot into the innovative applications being deployed that leverage AI technologies�. Out of these six papers, five deployed and one emerging application papers were selected. Deployed applications are �systems that have been in use for at least several months by individuals or organizations other than their developers.� Emerging applications are �technologies and systems that are close to deployment and clearly show�AI technologies.�

The deployed application topics dealt with load layout for ships, land management, assistance for online sales, synthetic agents for military training, and automation of space mission operations. They were titled as follows:

� RADARSAT-MAMM Automated Mission Planner � Interchanging Agents and Humans in Military Simulation � CARMA: A Case-Based Rangeland Management Adviser � NATURAL LANGUAGE ASSISTANT--A Dialog System for Online Product Recommendation � TALPS: The T-AVB Automated Load Planning System

The first of these listed projects, created by Benjamin Smith, Barbara Engelhardt, and Darren Mutz, describes the deployment of the �Automated Scheduling and Planning Environment� to �automate mission planning� in fall 2000. This deployment is significant because automating mission planning for this mission reduced work time from 52 work weeks to 8 work weeks and also enabled generation of many contingency plans. As a result, the generated plans performed flawlessly during the mission.

The �Interchanging Agents and Humans in Military Simulation� project by Clinton Heinze, Simon Goss, Torgny Josefson, Kerry Bennett, Sam Waugh, Ian Lloyd, Graeme Murray, and John Oldfield describes �the use of agent technology to assist in training for military operations.� This project allows humans to replace software and practice operations in challenging, real life scenarios. The next project, CARMA, is similar to this in the way that it supports decision making. It does so through �temporal projection, feature adaptation, and critical period adaptation� to maintain pest management and focus on forage consumption by grasshoppers.

The �Natural Language Assistant� project by Joyce Chai, Veronika Horvath, Nicolas Nicolov, Margo Stys-Budzikowska, Nanda Kambhatla, and Wlodek Zadrozny, further refines online decisions by offering �a system to help users find product and service information in electronic-commerce sites�. THE NLA �uses natural language processing to derive concepts of interest to the user� and provides a �concept hierarchy knowledge base to direct the user to the appropriate information.� The NLA boosts AI and electronic commerce. The last deployed application discussed was �TALPS: The T-AVB Automated Load Planning System� by Paul Cerkez. This paper describes rules and constraints to the �layout of military loads in T-AVB rapid-deployment ships� and uses information about a specific ship to rapidly develop a load plan. This system is currently in use by U.S. Marine Corps Aviation.

The emerging application paper explains how software agents can be used to �facilitate meetings, track individuals, gather and disseminate information, and assist other organizational efforts.� The authors of this project, Hans Chalupsky, Yolanda Gil, Craig Knoblock, Kristina Lerman, Jean Oh, David Pynadath, Thomas Russ, and Milind Tambe, titled the paper �ELECTRIC ELVES: Agent Technology for Supporting Human Organizations� and incorporate certain Al technologies, including �multi-agent and teamwork technologies to coordinate the many assets, information integration agents to find and assimilate many information sources, and knowledge representation and reasoning techniques to describe the agents and the knowledge they use.�

These six articles indicate how AI is more and more likely to enhance the future and the lives of people around the world. These applications show solutions to some of the more drastic endeavors and current problems that people will eventually have to face whether these issues are about insects, organizational efforts, mission flights, military simulations, rapid-deployment military ships, or commerce sites. As Steve Chien declares, �the AI community can take pride in these innovative applications and look ahead to even more fascinating and effective uses of AI technology in the future.�

Chien, Steve. Artificial Intelligence Magazine
June 22, 2002. �The Fourteenth Innovative Applications of Artificial Intelligence Conference� pgs. 1-3.
http://www.findarticles.com/cf_0/m2483/2_23/88702812/p1/article.jhtml?term=Artificial+Intelligence

1