Abstract
A methodology to rapidly produce environments that combine the intuition of in situ augmented reality (AR) with the commodity of virtual reality (VR) is proposed in this paper, by bringing together unmanned aerial systems (UAS) imagery and procedural modelling. While fully synthesized environments provide a very accurate visualization of the conserved parts of the real-world, missing parts - namely ruins - can be complemented with procedurally modelled structures. Regarding methodology's steps, firstly, a UAS flight mission gathers georeferenced imagery data about the site of interest. Then, the image set is converted to an accurate 3D model of the referred site, through photogrammetry. By considering the geographic information that also results from the previous process, ruins are manually outlined for georeferencing purposes. To complement ruins' missing information, virtual models of buildings are produced too, in a procedural modelling tool. Finally, at the full VR environment setup step, all elements are imported and subjected to geometric transformations that aim to match the procedurally modelled buildings with the outlined ruins. To improve the insight about the process work-flow, system's architecture and implementation are presented along with a case-study regarding a historically relevant site - Vila Velha's city gates (Vila Real, Portugal) - and preliminary results.

Abstract
Adaptability and energy-efficient sensing are essential properties to sustain the easy deployment and lifetime of WSNs. These properties assume a stronger role in autonomous sensing environments where the application objectives or the parameters under measurement vary, and human intervention is not viable. In this context, this paper proposes LiteSense, a self-adaptive sampling scheme for WSNs, which aims at capturing accurately the behavior of the physical parameters of interest in each WSN context yet reducing the overhead in terms of sensing events and, consequently, the energy consumption. For this purpose, a set of low-complexity rules auto-regulates the sensing frequency depending on the observed parameter variation. Resorting to real environmental datasets, we provide statistical results showing the ability of LiteSense in reducing sensing activity and power consumption, while keeping the estimation accuracy of sensing events.

Abstract
The growing trend of variable energy source integration in power systems (especially at a distribution level) is leading to an increased need for flexibility in all levels of the energy flows in such systems: the supply, the network and the demand sides. This paper focuses on a viable flexibility option that can be provided by means of a dynamic network reconfiguration (DNR), an automatic changing of line statuses in response to operational conditions in the system. The ultimate aim is to assess the impacts of such flexibility on the utilization levels of variable power sources (mainly, solar and wind) integrated at a distribution level. To perform this analysis, a stochastic mixed integer linear programming (S-MILP) operational model is developed in this work. The objective of the optimization problem is to minimize the sum of the most relevant cost terms while meeting a number of model constraints. The proposed model dynamically finds an optimal configuration of an existing network system in accordance with the system's operational conditions. The operation scale in the current work is one day, but with the possibility of an hourly reconfiguration. The standard IEEE 41-bus system is employed to test the proposed model and perform the analysis. Numerical results generally show that DNR leads to a more efficient utilization of renewable type DGs integrated in the system, reduced costs and losses, and a substantially improved system performance especially the voltage profile in the system.

Abstract
To achieve high availability in the face of network partitions, many distributed databases adopt eventual consistency, allow temporary conflicts due to concurrent writes, and use some form of per-key logical clock to detect and resolve such conflicts. Furthermore, nodes synchronize periodically to ensure replica convergence in a process called anti-entropy, normally using Merkle Trees. We present the design of Dotted-DB, a Dynamo-like key-value store, which uses a novel node-wide logical clock framework, overcoming three fundamental limitations of the state of the art: (1) minimize the metadata per key necessary to track causality, avoiding its growth even in the face of node churn; (2) correctly and durably delete keys, with no need for tombstones; (3) offer a lightweight antientropy mechanism to converge replicated data, avoiding the need for Merkle Trees. We evaluate DottedDB against MerkleDB, an otherwise identical database, but using per-key logical clocks and Merkle Trees for anti-entropy, to precisely measure the impact of the novel approach. Results show that: causality metadata per object always converges rapidly to only one id-counter pair; distributed deletes are correctly achieved without global coordination and with constant metadata; divergent nodes are synchronized faster, with less memory-footprint and with less communication overhead than using Merkle Trees.

Abstract
This paper proposes a new model to allocate reserve costs among the involved players, considering the characteristics of the several entities, and the particular circumstances at each moment. The proposed model is integrated in the Multi-Agent Simulator of Competitive Electricity Markets (MASCEM), which enables complementing the multi-agent simulation of diverse electricity market models, by including the joint simulation of energy and reserve markets. In this context, the proposed model allows allocating the payment of reserve costs that result from the reserve market. A simulation based on real data from the Iberian electricity market - MIBEL, is presented. Simulation results show the advantages of the proposed model in sharing the reserve costs fairly and accordingly to the different circumstances. This work thus contributes the study of novel market models towards the evolution of power and energy systems by adapting current models to the new paradigm of high penetration of renewable energy generation.

Abstract
Developing robust and universal methods for unsupervised segmentation of moving objects in video sequences has proved to be a hard and challenging task. The best solutions are, in general, computationally heavy preventing their use in real-time applications. This research addresses this problem by proposing a robust and computationally efficient method, BMOG, that significantly boosts the performance of the widely used MOG2 method. The complexity of BMOG is kept low, proving its suitability for real-time applications. The proposed solution explores a novel classification mechanism that combines color space discrimination capabilities with hysteresis and a dynamic learning rate for background model update.