The goal of representing context in a mixed initiative sys-tem is to model the information at a level of abstraction that is actionable for both the human and automated system. A potential solution to this problem is the Context for Human and Automation Teams (CHAT). This paper introduces the CHAT model and provides example implementations from several different applications such as task scheduling tech-niques, multi-agent systems, and human-robot interaction.

The focus of my research is an agent-based system for optimizing spatial arrangements of plants on a landscape to maximize their growth and minimize their water use. The optimization criteria include a natural phenomenon known as facilitation, which is observed in water-scarce environments when larger shrubs serve as benefactors to smaller annuals by generating conditions that protect them from harsh afternoon sun. In my modeling and optimization system each plant is an agent with growth requirements. A plant agent's fitness at a given location is defined by a fitness function that includes those growth requirements and a penalty term designed to force facilitation. The landscape design is formulated as a combinatorial optimization problem with a discrete set of locations for each plant on a grid, a fixed number of plants, and a fitness function that defines the performance of a plant at a location. To evaluate the effectiveness of this approach, I applied a variety of search strategies, including simulated annealing and a new agent-based approach that mimics how plant communities evolve over time, to different collections of simulated plant types and landscapes and compared the fitness scores and spatial arrangments in the solutions. The fitness scores from the search strategies were comparable. The search strategies produced different spatial distributions of the larger plants, and all designs exhibited facilitation and lower water use.

Cooperative path-finding can be abstracted as computing non-colliding paths for multiple agents between their start and goal locations on a graph. This work proposes a fast algorithm that can provide completeness guarantees for a general class of problems without any assumptions about the graph's topology. Specifically, the approach can address any solvable instance where there are at most n-2 agents in a graph of size n. The algorithm employs two primitives: a "push" operation where agents move towards their goals up to the point that no progress can be made, and a "swap" operation that allows two agents to swap positions without altering the configuration of other agents. Simulated experiments are provided on hard instances of cooperative path-finding, including comparisons against alternative methods. The results are favorable for the proposed algorithm and show that the technique scales to problems that require high levels of coordination, involving hundreds of agents.

We consider settings in which a semi-autonomous agent has uncertain knowledge about its environment, but can ask what action the human operator would prefer taking in the current or in a potential future state. Asking queries can improve behavior, but if queries come at a cost (e.g., due to limited operator attention), the value of each query should be maximized. We compare two strategies for selecting action queries: 1) based on myopically maximizing expected gain in long-term value, and 2) based on myopically minimizing uncertainty in the agent's policy representation. We show empirically that the first strategy tends to select more valuable queries, and that a hybrid method can outperform either method alone in settings with limited computation.

Human error is one of the most common causes of vulnerability in asecure system. However it is often overlooked when these systems aretested, partly because human tests are costly and very hard torepeat. We have developed a community of agents that test securesystems by running standard windows software while performingcollaborative group tasks, mimicking more realistic patterns ofcommunication and traffic, as well as human fatigue and errors. Thissystem is being deployed on a large cyber testing range. One keyattribute of humans is flexibility of response in order to achievetheir goals when unexpected events occur. Our agents use reactiveplanning within a BDI architecture to flexibly re-plan if needed.Since the agents are goal-oriented, we are able to measure the impactof cyber attacks on mission accomplishment, a more salient measure ofprotection than raw penetration. We show experimentally how the agentteams can be resilient under attacks that are partly successful, andalso how an organizational structure can lead to emergent propertiesof the traffic in the network.

Learning by demonstration technology has long held the promise to empower non-programmers to customize and extend software. We describe the deployment of a learning by demonstration capability to support user creation of automated procedures in a collaborative planning environment that is used widely by the U.S. Army. This technology, which has been in operational use since the summer of 2010, has helped to reduce user workloads by automating repetitive and time-consuming tasks. The technology has also provided the unexpected benefit of enabling standardization of products and processes. 

Robots with many sensors are capable of generating volumes of high-dimensional perceptual data. Making sense of this data and extracting useful knowledge from it is a difficult problem. For robots lacking proper models, trying to understand a stream of uninterpreted data is an especially acute problem. One critical step in linking raw uninterpreted perceptual data to cognition is dimensionality reduction. Current methods for reducing the dimension of data do not meet the demands of a robot situated in the world, and methods that use only perceptual data do not take full advantage of the interactive experience of an embodied robot agent. This work proposes a new scalable, incremental and active approach to dimensionality reduction suitable for extracting geometric knowledge from uninterpreted sensors and effectors. The proposed method uses distinctive state abstractions to organize early sensorimotor experience and sensorimotor embedding to incrementally learn accurate geometric representations based on experience. This approach is applied to the problem of learning the geometry of sensors, space, and objects. The result is evaluated using techniques from statistical shape analysis.

Most real-world negotiation involves multiple interdependent issues, which makes an agent's utility functions complex. Traditional negotiation mechanisms, which were designed for linear utilities, do not fare well in nonlinear contexts. One of the main challenges in developing effective nonlinear negotiation protocols is scalability; it can be extremely difficult to find high-quality solutions when there are many issues, due to computational intractability. One reasonable approach to reducing computational cost, while maintaining good quality outcomes, is to decompose the contract space into several largely independent sub-spaces. In this paper, we propose a method for decomposing a contract space into sub-spaces based on the agent's utility functions. A mediator finds sub-contracts in each sub-space based on votes from the agents, and combines the sub-contracts to produce the final agreement. We demonstrate, experimentally, that our protocol allows high-optimality outcomes with greater scalability than previous efforts. We also address incentive compatibility issues. Any voting scheme introduces the potential for strategic non-truthful voting by the agents, and our method is no exception. For example, one of the agents may always vote truthfully, while the other exaggerates so that its votes are always "strong." It has been shown that this biases the negotiation outcomes to favor the exaggerator, at the cost of reduced social welfare. We employ the limitation of strong votes to the method of decomposing the contract space into several largely independent sub-spaces. We investigate whether and how this approach can be applied to the method of decomposing a contract space.

A general game player automatically learns to play arbitrary new games solely by being told their rules. For this purpose games are specified in the game description language GDL, a variant of Datalog with function symbols and a few known keywords. In its latest version GDL allows to describe nondeterministic games with any number of players who may have imperfect, asymmetric information. We analyse the epistemic structure and expressiveness of this language in terms of epistemic modal logic and present two main results: The operational semantics of GDL entails that the situation at any stage of a game can be characterised by a multi-agent epistemic (i.e., S5-) model; (2) GDL is sufficiently expressive to model any situation that can be described by a (finite) multi-agent epistemic model.

A key problem in mechanism design is the construction of protocols that reach socially efficient decisions with minimal information revelation. This can reduce agent communication, and further, potentially increase privacy in the sense that agents reveal no more private information than is needed to determine an optimal outcome. This is not always possible: previous work has explored the tradeoff between communication cost and efficiency, and more recently, communication and privacy. We explore a third dimension: the tradeoff between privacy and efficiency. By sacrificing efficiency, we can improve the privacy of a variety of existing mechanisms. We analyze these tradeoffs in both second-price auctions and facility location problems (introducing new incremental mechanisms for facility location along the way). Our results show that sacrifices in efficiency can provide gains in privacy (and communication), in both the average and worst case.