Hand gestures offer a rich source of information for communication of spatial concepts. However, gestures are often lost when people work together in virtual space. Gestures may be conveyed in virtual spaces through many approaches, for example by drawing circles and arrows around objects on a shared screen, or using natural, real-time videos of a distant collaborator's hands or body. The first is an example of “surrogate” gestures and the second of “natural” gestures. The first is relatively inexpensive but requires users take some effort to draw and erase temporary symbols; the second is more costly but also more expressive and natural. This work will: 1) provide a brief, high-level framework characterizing tools that support virtual gestures; and 2) present an experimental study comparing the cognitive effort required when using sketched surrogate gestures versus natural gestures in a virtual mechanical design tasks. The larger goal is to understand whether the additional cost to support natural gestures is warranted, thus informing the design of future virtual collaboration tools. We recruited 24 participants and asked them to collaborate using three different approaches: face-to-face (as a baseline), virtual-sketching (which allows users to make gestural sketches on their drawings), and virtual-sketching with natural hand videos projected on the drawing surface. We found that users' cognitive workload was significantly reduced when natural hand videos were added to a virtual-sketching environment. These results suggest that the ease and expressivity of natural gestures provide more benefits than simpler surrogate gestures for virtual mechanical design tasks.

Hand gestures are a rich source of information in communication, particularly in communication of spatial concepts which are fundamental to engineering design tasks. However, there are many ways to convey gestures. For example, cursors can be used make simple gestures, so are hand images really necessary, or are simpler approaches sufficient? The work described in this paper is an evaluation comparing collaboration for a spatial design/drawing when users collaborate by three different methods: face-to-face, virtual sketching, and virtual sketching with the hand gestures of the virtual collaborators projected on the drawing surface. We found that users' performance in the face-to-face collaboration was almost identical to virtual collaboration with hand images. Virtual collaboration without hand images was significantly more difficult. These results suggest that rich hand gestures conveyed with the hands in face-to-face settings are important in joint spatial tasks; their loss reduces effectiveness in virtual collaboration on such tasks, but the addition of hand images projected on a joint work surface can restore effectiveness to levels similar to that of face-to-face collaboration.

Time and space have always been constrains in providing health care. Telemedicine is changing the face of health care by increasing access to health care for patients in remote and underserved areas, which in turn enables them to manage their health better. This translates to cost savings for providers. However, clinics and hospitals often find it challenging to introduce telemedicine into their organizations for many reasons, one of which is the perception that large infrastructure investments must be made at the onset. This paper presents a brief history of telemedicine, followed by several examples telemedicine health care deliver approaches used in a major U.S. health provider, and several approaches for making flexible use of clinic space to support both distance and in-person patients. We provide these examples to help health care organizations find ways to enter into telemedicine in small ways before making large scale investments. Flexible systems revolutionized the manufacturing industry; we feel that similar approaches can revolutionize telemedicine.

This study compares benefits of two multi-criteria decision making techniques: Multi-criteria Rank Ordering (MRO) and Analytical Hierarchy Process (AHP). The study is intended to help software designers determine which method to implement when helping users compare alternatives. The AHP method requires more user time and effort than MRO; our goal was to explore whether AHP provided sufficient benefits over MRO to warrant the additional time it requires. Eighteen subjects were asked to use software tools implementing AHP and MRO. They compared student application files and identified applicants most qualified for admission to graduate school. The results showed that although AHP required more time than MRO, AHP resulted in lower variance in the assessments produced by different decision makers. This may be explained by the richer information collected by the AHP method, resulting in more accurate reflections of decision makers' assessments. Thus, for high criticality domains, AHP may be warranted.

A difficulty for military analysts is to thoroughly consider possible plans which the enemy might follow under various assumptions. However, it is critical to do so to avoid dangerous surprises. Weasel is a decision support system that assists intelligence analysts in developing alternative enemy courses of action (ECOAs) consistent with intelligence conclusions and specified assumptions. A design goal behind Weasel is to provide users with the convenience of automation while still providing them with flexibility and control over the results. Weasel allows users to perform "what if" analyses in which they can change assumptions about enemy resources, current position, and maneuver constraints. Weasel integrates two other decision support tools, CoRaven, which assists in analyzing intelligence data, and Fox, which assists in generating friendly courses of action (FCOAs). The combination of CoRaven, Weasel and Fox can provide powerful assistance in the military tactical decision making process.

Users at different levels of domain experience have very different needs. For example, a system designed to assist domain novices may frustrate experts and vice-versa. This is one of several challenges specific to building decision support systems for experience-centered domains. A second challenge in working with complex experience-centered domains is that it is hard for non-experts to understand the domain in order to model it. In this paper we present DAISY, the design aid for intelligent support systems. It is a software design methodology for constructing decision support systems in complex, experience-based domains. DAISY address the specialized challenges of these domains by augmenting existing cognitive engineering methodologies. In particular, DAISY provides a method for identifying the specialized needs of users within a specific range of domain experience. Thus, it can help software designers to understand "What does the domain expert need?" or "What does a trained novice need?" To help system designers manage the complexity of modeling unfamiliar experience-centered domains, it provides a tool called a time/activity matrix. To illustrate each of DAISY's steps, we used the development of a decision support system called Fox. Fox assists expert military planners by rapidly generating alternative plans. This is a cognitively difficult, time critical task with life and death consequences.

A challenge for the designers of decision support systems (DSSs) is that the introduction of a decision aid into a complex domain often generates new, unexpected user needs. In this paper, we discuss how the iterative application of an experience-centered design methodology called DAISY (Design Aid for Intelligent Support SYstems) provides concrete methods for modeling these changes and identifying new system requirements caused by the introduction of the DSS. We illustrate the iterative use of DAISY in the design, evaluation and modification of Fox, a DSS intended to assist expert military users by helping them to generate and evaluate a broad range of plan options.

Describes a model-based design method to develop CoRaven, a decision support tool that is intended to assist military intelligence analysts in managing and interpreting large quantities of battlefield information. In this method, we use observations of practitioners solving specific tasks in order to understand and model how they use information. We use this model of the task to help identify user needs that the tool must support, and, during initial prototyping, to guide usability analyses. We have found task models to be an important consideration in the decision support tool design process that can help to constrain the design space and reduce the time required to develop an effective decision support tool prototype.

Agent-based approaches are becoming increasingly important because of their generality, flexibility, modularity, and ability to take advantage of distributed resources. Agents are used in information retrieval, entertainment, coordinating multiple robots, and modeling economic systems. Agents can recommend music, tell stories, and interact with people. They are useful for reducing humans' work and information load in tasks such as medical monitoring and battlefield reasoning. Agents are already changing the way in which we gather information, manage investments, and conduct business. This article provides an introduction to agent issues, outlines motivations for using agent-based paradigms, and describes some of their current uses.

This paper presents Coordinator, a robust, resource adaptive process planner for CNC machining centers that uses a blend of geometric and manufacturing information to detect a broad variety of manufacturing interactions and create setups plans. This system is designed to be used by expert users to reduce the time and effort required to produce good plans. One of Coordinator's contributions is the ability to detect fixture-feature interactions, interferences between manufacturing operations in which one operation destroys the clamping surfaces required by another. These are an important class of interactions, which have a major (if not the major) impact on ordering setup plans. However, they are rarely addressed in the literature, mainly because traditional process planning organizations do not compute the information required to detect them at the right time. Fixturing information, which is needed for identifying feature-fixture interactions, is usually computed after all plan sequencing has been done. Coordinator is part of a large system, MAPP, which reorganized the tasks in process planning by computing the required abstract fixturing information prior to sequencing so that Coordinator can detect a wide variety of interaction types, including feature-fixture interactions. Coordinator represents an important advance in making practical, robust, and flexible manufacturing planners.

Designing systems to effectively assist planners in grasping the situation quickly and making high quality decisions is very difficult, even within a single problem solving domain. Different types of users have very different needs, and a system designed to assist one group of users may frustrate other users with differing amounts of experience. We present DAISY, a methodology for developing planning aids for users, given a specified level of expertise. This methodology is intended to enable system designers to identify the system requirements needed to meet the information and display needs of users at a given level of experience prior to designing the system. These requirements are identified through user problem solving studies and the development of a model of the task, the information requirements and the typical user errors. The DAISY methodology is unique in that it identifies the needs of planners at varying ranges of experience, and provides concrete methods for incorporating these specialized user needs into the software design. We illustrate the use of this methodology in the design of an intelligent agent and human-computer interface, Fox, for the military planning task of course of action generation. This is a complex and difficult decision making task in which users make life and death decisions while they are are under extreme time pressure and overloaded with information.

Describes FOX-GA, an intelligent planning decision support tool for assisting military intelligence and maneuver battlestaff in rapidly generating and assessing battlefield courses of action (COAs). The motivations behind FOX stem from the need to plan and replan rapidly in such a way as to allow users flexibility and control over planning objectives and options. The environment in which plans are executed (the battlefield) is one that is inherently uncertain and rapidly changing, demanding frequent re-planning during execution. To help meet these rapid replanning needs, we designed FOX to rapidly generate and evaluate a broader variety of variety of high quality COAs faster than they could do so themselves. FOX then evaluates the COAs and presents only the best few to users, and allows them to reassess those options according to their own judgment, and to either edit or select the ones they feel are best. Early evaluations indicate that users explore a wider variety of COAs with FOX then without.

Intelligence analysis is one of the major functions performed by an Army staff in battlefield management. In particular, intelligence analysts develop intelligence requirements based on the commander's information requirements, develop a collection plan, and then monitor messages from the battlefield with respect to the commander's information requirements. The goal of the CoRAVEN project is to develop an intelligent collaborative multimedia system to support intelligence analysts. Key ingredients of our design approach include: (1) significant knowledge engineering activities with domain experts, (2) representation of an explicit model of reasoning and activity to drive design, (3) the use of Bayesian belief networks as a way to structure inferences that relate observable data to the commander's information requirements, (4) collaborative graphical user interfaces to provide flexible support for the multiple tasks in which analysts are engaged, (5) sonification of data streams and alarms to support enhanced situation awareness, (6) detailed psychological studies of reasoning and judgment under uncertainty, and (7) iterative prototyping of candidate designs with domain experts for both formative and summative evaluation. The paper discusses our current progress on all these fronts.

This paper describes the Quality and Experience Metric (QUEM), a method for estimating the skill level of a knowledge based system based on the quality of the solutions it produces. It allows one to assess how many years of experience the system would be judged to have if it were a human by providing a quantitative measure of the system's overall competence. QUEM can be viewed as a type of achievement or job placement test administered to knowledge based systems to help system designers determine how the system should be used and by what level of user. To apply QUEM, a set of subjects, experienced judges, and problems must be identified. The subjects should have a broad range of experience levels. Subjects and the knowledge based system are asked to solve the problems; and judges are asked to rank order all solutions, from worst quality to best. The data from the subjects is used to construct a skill function relating experience to solution quality, and confidence bands showing the variability in performance. The system's quality ranking is then plugged into the skill function to produce an estimate of the system's experience level. QUEM can be used to gauge the experience level of an individual system, to compare two systems, or to compare a system to its intended users. This represents an important advance in providing quantitative measures of overall performance that can be applied to a broad range of systems.

In complex practical domains, such as manufacturing, plans generated must not only be feasible, they must also be of high quality (i.e., efficient, reliable, etc.). The particular quality criterion used is determined according to the objectives and needs of the planner. This paper describes P/sup 3/, an automated process planner for generating manufacturing plans for prismatic machining processes. P/sup 3/ is constraint-based, and it utilizes the LCOS (least commitment to operator selection strategy) to facilitate the construction of efficient, high quality plans. The ability to develop high quality plans depends on the ability to make trade-offs at a global level among factors that are frequently conflicting. LCOS is a search strategy that facilitates these objectives by making the appropriate information available so as to allow globally efficient plans to be constructed. P/sup 3/ implements the LCOS strategy in an algorithm called MACHINE, which combines LCOS with a variety of search pruning strategies to construct globally efficient plans while avoiding rapid combinatorial explosion. Lastly, P/sup 3/'s operation is illustrated through the construction of an illustrated example manufacturing plan, and its performance will be evaluated. The result is a planner which is more scalable to problems containing large numbers of goals, while still maintaining high solution quality. It is hoped that the general strategies presented can provide useful and scalable frameworks for modeling planning problems in other complex practical domains.