Emerging wearable sensors have enabled the unprecedented ability to continuously monitor human activities for healthcare purposes. However, with so many ambient sensors collecting different measurements, it becomes important not only to maintain good monitoring accuracy, but also low power consumption to ensure sustainable monitoring. This power-efficient sensing scheme can be achieved by deciding which group of sensors to use at a given time, requiring an accurate characterization of the trade-off between sensor energy usage and the uncertainty in ignoring certain sensor signals while monitoring. To address this challenge in the context of activity monitoring, we have designed an adaptive activity monitoring framework. We first propose a switching Gaussian process to model the observed sensor signals emitting from the underlying activity states. To efficiently compute the Gaussian process model likelihood and quantify the context prediction uncertainty, we propose a block circulant embedding technique and use Fast Fourier Transforms (FFT) for inference. By computing the Bayesian loss function tailored to switching Gaussian processes, an adaptive monitoring procedure is developed to select features from available sensors that optimize the trade-off between sensor power consumption and the prediction performance quantified by state prediction entropy. We demonstrate the effectiveness of our framework on the popular benchmark of UCI Human Activity Recognition using Smartphones.

Years ago when I worked in the semiconductor crystal growth industry, we had a frequent problem with crystal defects that occurred, seemingly at random, during a multi-day growth process. These defects caused total failure of the greater than $80,000 single crystal. One of my proposed solutions was to create a dynamic computer model of the whole system. The simulation would run simultaneously with the real system's control settings and model in real-time the various potentials for defects well in advance of their occurring. The closed-loop SCADA [1] system would then make appropriate corrections to avoid defects and yield a perfect crystal.

Remember Innerspace, the comedy sci-fi movie from the '80s about a microscopic manned pod injected into a human? Although we're years away from launching submarines inside our bodies, advances in engineering have made it possible to build computers so tiny that embedding them inside living tissue is no longer a figment of a sci-fi writer's imagination. Indeed, it's now been 20 years since British scientist Kevin Warwick first implanted a silicon RFID transmitter into his arm to remotely control computers in doors, lights and other devices. He then took it a step further by interfacing the device with his own nervous system to control a robotic arm, earning himself the nickname "Captain Cyborg." While it's not headline news every day, the pace of microcomputer technology has not slowed, and I'm still occasionally astounded by the ingenuity of some new developments.

The NIMBUS Lab at the University of Nebraska has been developing drones that have the unique ability to dig holes in the ground and then fill those holes with sensors. If this sounds like a complicated task, that's because it is: The drone needs to be able to carry a portable digging system a useful distance, locate a diggable spot, land, verify that the spot it thought was diggable is in fact diggable, dig a hole and install the sensor, and then fly off again. At IROS late last year, folks from the NIMBUS Lab presented a paper detailing a rather burly quadcopter that could carry an auger with an embedded sensor and use it to place the sensor in the ground (you can see a video of this in action here). And at ISER a few weeks later, they presented another paper on how the drone can autonomously figure out whether it's digging in a good spot or not. One of the biggest challenges to a system like this is that by the time you pack in the drilling rig and all the sensors and computers that the drone needs to operate autonomously, you'll be lucky if the thing will manage to keep itself aloft for more than just a few minutes.

LAS VEGAS--Get ready to peek at a new Ring. For homes and apartments that don't have doorbells, the folks behind the Ring video doorbell have a new offering, the Ring Door View Cam. The company describes it as "a wire-free video doorbell that transforms a door viewer into a smart security device." Jamie Siminoff, the founder of Ring, says the peep hole cam has been one of the top requests from customers. "There's a small subset of people who don't have room for a door bell," he says.

We present an efficient convolution kernel for Convolutional Neural Networks (CNNs) on unstructured grids using parameterized differential operators while focusing on spherical signals such as panorama images or planetary signals. To this end, we replace conventional convolution kernels with linear combinations of differential operators that are weighted by learnable parameters. Differential operators can be efficiently estimated on unstructured grids using one-ring neighbors, and learnable parameters can be optimized through standard back-propagation. As a result, we obtain extremely efficient neural networks that match or outperform state-of-the-art network architectures in terms of performance but with a significantly lower number of network parameters. We evaluate our algorithm in an extensive series of experiments on a variety of computer vision and climate science tasks, including shape classification, climate pattern segmentation, and omnidirectional image semantic segmentation. Overall, we present (1) a novel CNN approach on unstructured grids using parameterized differential operators for spherical signals, and (2) we show that our unique kernel parameterization allows our model to achieve the same or higher accuracy with significantly fewer network parameters.

Bio-hybrids providing new energy sources, AI helping shoppers to try before they buy, currencies becoming a free-for-all and algorithmically-driven peak journey planners helping train operators avoid congestion are just some of the future forces that will shape 2018. A key overarching theme is digital decentralisation: where technology is empowering brands and communities to seize the initiative and challenge centralised ways of doing things. This trend is being seen across all sectors but specifically in finance, in the form of cryptocurrencies and energy with the rise of micro-grids. Trends are less about forecasting the future than reshaping the present by unlocking possibilities and helping businesses to deepen their understanding of what is achievable. In 2018 the most interesting techno-societal changes are happening in energy, retail, mobility and money.

Microscopic (pore-scale) properties of porous media affect and often determine their macroscopic (continuum- or Darcy-scale) counterparts. Understanding the relationship between processes on these two scales is essential to both the derivation of macroscopic models of, e.g., transport phenomena in natural porous media, and the design of novel materials, e.g., for energy storage. Most microscopic properties exhibit complex statistical correlations and geometric constraints, which presents challenges for the estimation of macroscopic quantities of interest (QoIs), e.g., in the context of global sensitivity analysis (GSA) of macroscopic QoIs with respect to microscopic material properties. We present a systematic way of building correlations into stochastic multiscale models through Bayesian networks. This allows us to construct the joint probability density function (PDF) of model parameters through causal relationships that emulate engineering processes, e.g., the design of hierarchical nanoporous materials. Such PDFs also serve as input for the forward propagation of parametric uncertainty; our findings indicate that the inclusion of causal relationships impacts predictions of macroscopic QoIs. To assess the impact of correlations and causal relationships between microscopic parameters on macroscopic material properties, we use a moment-independent GSA based on the differential mutual information. Our GSA accounts for the correlated inputs and complex non-Gaussian QoIs. The global sensitivity indices are used to rank the effect of uncertainty in microscopic parameters on macroscopic QoIs, to quantify the impact of causality on the multiscale model's predictions, and to provide physical interpretations of these results for hierarchical nanoporous materials.

Compressive sensing (CS) has been studied and applied in structural health monitoring for wireless data acquisition and transmission, structural modal identification, and spare damage identification. The key issue in CS is finding the optimal solution for sparse optimization. In the past years, many algorithms have been proposed in the field of applied mathematics. In this paper, we propose a machine-learning-based approach to solve the CS data-reconstruction problem. By treating a computation process as a data flow, the process of CS-based data reconstruction is formalized into a standard supervised-learning task. The prior knowledge, i.e., the basis matrix and the CS-sampled signals, are used as the input and the target of the network; the basis coefficient matrix is embedded as the parameters of a certain layer; the objective function of conventional compressive sensing is set as the loss function of the network. Regularized by l1-norm, these basis coefficients are optimized to reduce the error between the original CS-sampled signals and the masked reconstructed signals with a common optimization algorithm. Also, the proposed network can handle complex bases, such as a Fourier basis. Benefiting from the nature of a multi-neuron layer, multiple signal channels can be reconstructed simultaneously. Meanwhile, the disassembled use of a large-scale basis makes the method memory-efficient. A numerical example of multiple sinusoidal waves and an example of field-test wireless data from a suspension bridge are carried out to illustrate the data-reconstruction ability of the proposed approach. The results show that high reconstruction accuracy can be obtained by the machine learning-based approach. Also, the parameters of the network have clear meanings; the inference of the mapping between input and output is fully transparent, making the CS data reconstruction neural network interpretable.

THE importance of science in society has no greater spokesperson than Lord Martin Rees. From his perch at Cambridge--and a centre he formed on studying existential risks--he has served as both a promoter, populariser and the moral conscience of scientific endeavour far beyond his academic field of astrophysics. In "Our Final Century" in 2003 (retitled more breathlessly "Our Final Hour" in the American edition) he presented a range of global challenges, from bioterrorism to nuclear weapons. He put the risk of human extinction by 2100 from our technologies at around 50%. His latest book, "On the Future", is more sanguine.