Wilderness Search and Rescue ͑WiSAR͒ entails searching over large regions in often rugged remote areas. Because of the large regions and potentially limited mobility of ground searchers, WiSAR is an ideal application for using small ͑human-packable͒ unmanned aerial vehicles ͑UAVs͒ to provide aerial imagery of the search region. This paper presents a brief analysis of the WiSAR problem with emphasis on practical aspects of visual-based aerial search. As part of this analysis, we present and analyze a generalized contour search algorithm, and relate this search to existing coverage searches. Extending beyond laboratory analysis, lessons from field trials with search and rescue personnel indicated the immediate need to improve two aspects of UAV-enabled search: How video information is presented to searchers and how UAV technology is integrated into existing WiSAR teams. In response to the first need, three computer vision algorithms for improving video display presentation are compared; results indicate that constructing temporally localized image mosaics is more useful than stabilizing video imagery. In response to the second need, a goal-directed task analysis of the WiSAR domain was conducted and combined with field observations to identify operational paradigms and field tactics for coordinating the UAV operator, the payload operator, the mission manager, and ground searchers.

No abstract

Effective presentation of tumor antigens is fundamental to strategies aimed at enrolling the immune system in eradication of residual disease after conventional treatments. Myeloid malignancies provide a unique opportunity to derive dendritic cells (DCs), functioning antigenpresenting cells, from the malignant cells themselves. These may then co-express leukemic antigens together with appropriate secondary signals and be used to generate a specific, antileukemic immune response. In this study, blasts from 40 patients with acute myeloid leukemia (AML) were cultured with combinations of granulocyte-macrophage colony-stimulating factor, interleukin 4, and tumor necrosis factor ␣, and development to DCs was assessed. After culture, cells from 24 samples exhibited morphological and immunophenotypic features of DCs, including expression of major histocompatibility complex class II, CD1a, CD83, and CD86, and were potent stimulators in an allogeneic mixed lymphocyte reaction (MLR). Stimulation of autologous T-cell responses was assessed by the proliferative response of autologous T cells to the leukemic DCs and by demonstration of the induction of specific, autologous, antileukemic cytotoxicity. Of 17 samples, 11 were effective stimulators in the autologous MLR, and low, but consistent, autologous, antileukemic cytotoxicity was induced in 8 of 11 cases (mean, 27%; range, 17%-37%). This study indicates that cells with enhanced antigen-presenting ability can be generated from AML blasts, that these cells can effectively prime autologous cytotoxic T cells in vitro, and that they may be used as potential vaccines in the immunotherapy of AML. (Blood. 2001; 97:2764-2771)

Young-of-year piscivores typically undergo ontogenetic diet shifts from planktivory to benthivory to piscivory. These shifts are often the result of changes in predator foraging abilities, but little is known about the influence of relative prey availability. As a result, we examined diet shifts across a range of sizes (20–150 mm) of a young-of-year piscivore, walleye (Sander vitreus), in feeding experiments in which zooplankton, benthic invertebrates, and fish were made available at different density combinations. Consumption of each prey type changed with walleye size and prey densities. Small juveniles (20 mm) selected zooplankton and fish, whereas larger walleye (40–100 mm) selected benthic invertebrates and fish. Relative prey densities influenced consumption; increased densities of more profitable prey types resulted in reduced consumption of less profitable prey. Walleye larger than 100 mm selected only fish. Foraging efficiencies also varied with size and prey types; small walleye (20 mm) were less likely to pursue benthic invertebrates and retain captured fish. These ontogenetic changes in foraging patterns are linked to prey profitability, have growth consequences for juvenile walleye, and have implications for understanding diet shifts of other juvenile fish.

No abstract

This paper focuses on coalition formation for task allocation in both multiagent and multi-robot domains. Two different problem formalizations are considered, one for multi-agent domains where agent resources are transferable and one for multi-robot domains. We demonstrate complexity theoretic differences between both models and show that, under both, the coalition formation problem, with m tasks, is NP-hard to both solve exactly and to approximate within a factor of O(m 1− ) for all > 0. Two natural restrictions of the coalition formation problem are considered. In the first situation agents are drawn from a set of j types. Agents of each type are indistinguishable from one another. For this situation a dynamic programming based approach is presented, which, for fixed j finds the optimal coalition structure in polynomial time and is applicable in both multi-agent and multi-robot domains. We then consider situations where coalitions are restricted to k or fewer agents. We present two different algorithms. Each guarantees the generated solution to be within a constant factor, for fixed k, of the optimal in terms of utility. Our algorithms complement Shehory and Kraus' algorithm (Artif Intell 101(1-2): 1998), which provides guarantee's on solution cost, as ours provides guarantees on utility. Our algorithm for general multi-agent domains is a modification of and has the same running time as Shehory and Kraus' algorithm, while our approach for multi-robot domains runs in time O(n 3 2 m), much faster than Vig and Adams (J Intell Robot Syst 50(1): 2007) modifications to Shehory and Kraus' algorithm for multi-robot domains, which ran in time O(n k m), for n agents and m tasks.

No abstract

Converging evidence from psychology, human factors, management and organizational science, and other related fields suggests that humans working in teams employ shared mental models to represent and use pertinent information about the task, the equipment, the team members, and their roles. In particular, shared mental models are used to interact efficiently with other team members and to track progress in terms of goals, subgoals, achieved and planned states, as well as other team-related factors. Although much of the literature on shared mental models has focused on quantifying the success of teams that can use them effectively, there is little work on the types of data structures and processes that operate on them, which are required to operationalize shared mental models. This paper proposes the first comprehensive formal and computational framework based on results from human teams that can be used to implement shared mental models for artificial virtual and robotic agents. The formal portion of the framework specifies the necessary data structures and representations, whereas the computational framework specifies the necessary computational processes and their interactions to build, update, and maintain shared mental models.