Beyond the Elves: Making Intelligent Agents Intelligent

نویسنده

  • Craig A. Knoblock
چکیده

(Chalupsky et al. 2001, 2002) at the University of Southern California (USC) was to develop software agents to support human organizations. The project was quite successful with impressive prototypes and many papers on the research. In fact, DARPA, which funded the project, was so enthusiastic about the results that it asked us to deploy a version of the Elves for use at DARPA. The original application of the Elves deployed at USC was for the office environment (called the Office Elves) (Scerri, Pynadath, and Tambe 2002; Pynadath and Tambe 2003) and required detailed information about the calendars of people using the system. Deploying this application at DARPA raised a host of issues including privacy and access to internal services. Thus, we decided to deploy a new application of the Electric Elves, called the Travel Elves. This application appeared to be ideal for wider deployment since it could be hosted entirely outside an organization and communication could be performed over wireless devices, such as cellular telephones. The mission of the Travel Elves (Ambite et al. 2002, Knoblock 2004) was to facilitate planning a trip and to ensure that the resulting travel plan would execute smoothly. Initial deployment of the Travel Elves at DARPA went smoothly. Program managers and office directors began using the system. We trained the DARPA travel administrator on how to enter the travel information into the system. But over time things began to go wrong: agents would fail to correctly execute their assigned tasks, information got delayed or was not sent at all, the underlying sources were continually changing, and each user wanted the system to operate in different ways. In this paper, we first review the work on the Travel Elves and describe what the system did and how it worked. Then we describe the real-world problems that arose in deploying this application and the underlying causes of these failures. Finally, we will present some lessons learned and recent research that was motivated by our experiences in deploying the Elves at DARPA.

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Adjustable Autonomy: A Response

Gaining a fundamental understanding of adjustable autonomy (AA) is critical if we are to deploy multi-agent systems in support of critical human activities. Indeed, our recent work with intelligent agents in the “Electric Elves” (E-Elves) system has convinced us that AA is a critical part of any human collaboration software. In the following, we first briefly describe E-Elves, then discuss AA i...

متن کامل

Outsourcing or Insourcing of Transportation System Evaluation Using Intelligent Agents Approach

Nowadays, outsourcing is viewed as a trade strategy and organizations tend to adopt new strategies to achieve competitive advantages in the current world of business. focusing on main copmpetencies, and transferring most of activities to outside resources of organization( outsourcing) is one such strategy is. In this paper, we aim to decide on decision maker agent of transportation system, by a...

متن کامل

A Context-aware Architecture for Mental Model Sharing through Semantic Movement in Intelligent Agents

Recent studies in multi-agent systems are paying increasingly more attention to the paradigm of designing intelligent agents with human inspired concepts. One of the main cognitive concepts driving the core of many recent approaches in multi agent systems is shared mental models. In this paper, we propose an architecture for sharing mental models based on a new concept called semantic movement....

متن کامل

Simulation for Development and Assessment of Cyber Battlespace Intelligent Agents

he battlespace is increasingly dynamic. The volume of information available to all decision-makers is increasing. The uncertainties about the battlespace, whether due to cyber attack, miscommunication, mechanical fault, or other problems, prevents rapid, effective, informed decision-making. Clearly, the mere application of so-called “big data” technologies cannot address the scope of the coming...

متن کامل

User-based Vehicle Route Guidance in Urban Networks Based on Intelligent Multi Agents Systems and the ANT-Q Algorithm

Guiding vehicles to their destination under dynamic traffic conditions is an important topic in the field of Intelligent Transportation Systems (ITS). Nowadays, many complex systems can be controlled by using multi agent systems. Adaptation with the current condition is an important feature of the agents. In this research, formulation of dynamic guidance for vehicles has been investigated based...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:
  • AI Magazine

دوره 29  شماره 

صفحات  -

تاریخ انتشار 2006