Air pollution has a direct impact to human health, and data-driven air quality models are useful for evaluating population exposure to air pollutants. In this paper, we propose a novel region-based Gaussian process model for estimating urban air pollution dispersion, and applied it to a large dataset of ultrafine particle (UFP) measurements collected from a network of sensors located on several trams in the city of Zurich. We show that compared to existing grid-based models, the region-based model produces better predictions across aggregates of all time scales. The new model is appropriate for many useful user applications such as exposure assessment and anomaly detection.

Sensing and monitoring of our natural environment are important for sustainability. As sensor systems grow to a large scale, it will become infeasible to place all sensors under centralized control. We investigate community sensing, where sensors are controlled by self-interested agents that report their measurements to a center. The center can control the agents only through incentives that motivate them to provide the most accurate and useful reports. We consider different game-theoretic mechanisms that provide such incentives and analyze their properties. As an example, we consider an application of community sensing for monitoring air pollution.

Qualitative Spatial and Temporal Reasoning (QSTR) represents spatial and temporal information in terms of human comprehensible qualitative predicates and reasons about qualitative information by solving qualitative constraint networks (QCNs). Despite significant progress in the past three decades, more and more evidence has shown that it is inherently hard to find exact solutions for expressive qualitative constraints. In many applications, however, we are often required to make decisions in a very limited time. In these cases, finding a good approximate solution in seconds is much more desirable than waiting days for an exact solution. In this paper, we will exploit the algebraic structure of qualitative calculi (e.g. Interval Algebra and RCC8) as well as their conceptual neighbourhood graphs to develop approximate methods for consistency checking in QSTR. Moreover, we propose and empirically compare four independent methods to serve as tools for finding good approximate solutions for the given qualitative calculi.

Monitoring and managing urban air pollution is a significant challenge for the sustainability of our environment. We quickly survey the air pollution modeling problem,introduce a new dataset of mobile air quality measurements in Zurich, and discuss the challenges of making sense of these data.

Representing and reasoning spatial and temporal information is a key research issue in Computer Science and Artificial Intelligence. In this paper, we introduce tools that produce three novel encodings which translate problems in qualitative spatial and temporal reasoning into logic programs for answer set programming solvers. Each encoding reflects a different type of modeling abstraction. We evaluate our approach with two of the most well known qualitative spatial and temporal reasoning formalisms, the Interval Algebra and Region Connection Calculus. Our results show some surprising findings, including the strong performance of the solver for disjunctive logic programs over the non-disjunctive ones on our benchmark problems.

The next challenge in qualitative spatial and temporal reasoning is to develop calculi that deal with different aspects of space and time. One approach to achieve this is to combine existing calculi that cover the different aspects. This, however, can lead to calculi that have a very large number of relations and it is a matter of ongoing discussions within the research community whether such large calculi are too large to be useful. In this paper we develop a procedure for reasoning about some of the largest known calculi, the Rectangle Algebra and the Block Algebra with about 10661 relations. We demonstrate that reasoning over these calculi is possible and can be done efficiently in many cases. This is a clear indication that one of the main goals of the field can be achieved: highly expressive spatial and temporal representations that support efficient reasoning.

This paper introduces restart and nogood recording techniques in the domain of qualitative spatial and temporal reasoning. Nogoods and restarts can be applied orthogonally to usual methods for solving qualitative constraint satisfaction problems. In particular, we propose a more general definition of nogoods that allows for exploiting information about nogoods and tractable subclasses during backtracking search. First evaluations of the proposed techniques show promising results.

Constraint networks in qualitative spatial and temporal reasoning are always complete graphs. When one adds an extra element to a given network, previously unknown constraints are derived by intersections and compositions of other constraints, and this may introduce inconsistency to the overall network. Likewise, when combining two consistent networks that share a common part, the combined network may become inconsistent.