Recently, a hybrid Deep Neural Network (DNN) algorithm, TreNet was proposed for predicting trends in time series data. While TreNet was shown to have superior performance for trend prediction to other DNN and traditional ML approaches, the validation method used did not take into account the sequential nature of time series data sets and did not deal with model update. In this research we replicated the TreNet experiments on the same data sets using a walk-forward validation method and tested our optimal model over multiple independent runs to evaluate model stability. We compared the performance of the hybrid TreNet algorithm, on four data sets to vanilla DNN algorithms that take in point data, and also to traditional ML algorithms. We found that in general TreNet still performs better than the vanilla DNN models, but not on all data sets as reported in the original TreNet study. This study highlights the importance of using an appropriate validation method and evaluating model stability for evaluating and developing machine learning models for trend prediction in time series data.

Reinforcement Learning (RL) is an effective tool for controller design but can struggle with issues of robustness, failing catastrophically when the underlying system dynamics are perturbed. The Robust RL formulation tackles this by adding worst-case adversarial noise to the dynamics and constructing the noise distribution as the solution to a zero-sum minimax game. However, existing work on learning solutions to the Robust RL formulation has primarily focused on training a single RL agent against a single adversary. In this work, we demonstrate that using a single adversary does not consistently yield robustness to dynamics variations under standard parametrizations of the adversary; the resulting policy is highly exploitable by new adversaries. We propose a population-based augmentation to the Robust RL formulation in which we randomly initialize a population of adversaries and sample from the population uniformly during training. We empirically validate across robotics benchmarks that the use of an adversarial population results in a more robust policy that also improves out-of-distribution generalization. Finally, we demonstrate that this approach provides comparable robustness and generalization as domain randomization on these benchmarks while avoiding a ubiquitous domain randomization failure mode.

Digital transformation is one of the top priorities for industrial companies. The largest players are already moving in this direction, for many years continuously working to improve production efficiency and launching large-scale optimisation programs. They're called advanced analytics or digital innovation, and at their core, the technology could be summarised under artificial intelligence. In all cases, the efforts to utilise AI models or data analytics systems are part of a bigger digital transformation effort of the progressing companies. In an industrial context, such strategies for cost-saving and process optimisation often start from pilot projects, or top management directives for digital change guide them. In general, changes in processes or investments in capital-intensive and competitive industries require large sums of money. Traditional capital expenditures usually stretch over a long period, so a current financial standing may not allow for a complete physical overhaul of the plants or facilities. These high costs lead to the search for cheaper alternatives.

The UK's HS2 has trialled a new artificial intelligence-based carbon and cost estimating solution to decrease carbon emissions. The technology was trialled at several HS2 locations managed by the Skanska Costain STRABAG joint venture. The solution helps in automating building information model (BIM) processes. It enables simulating multiple design options using different combinations and types of construction materials. The process helps in measuring and comparing the environmental impacts and carbon emissions for each simulation and accordingly design a cost-effective and environmentally friendly construction model.

Algorithms have taken a lot of flak recently, particularly those being used by the government and other public bodies in the UK. The controversial algorithm used to award student grades caused a huge public outcry, but national and local governments and several police forces have been withdrawing other algorithms and artificial intelligence tools from use throughout the year in response to legal challenges and design failures. This has quite rightly brought it home to public sector organisations that a more critical approach to AI and algorithmic decision-making is needed. But there are many cases in which government bodies can deploy such technology in lower risk, high-impact scenarios that can improve lives, particularly if they don't directly use personal data. So before we leap full pelt into AI cynicism we should consider benefits as well as risks it offers, and demand a more responsible approach to AI development and deployment.

A major goal of material design is the inverse optimization of processing-structure-property relationships. In this paper, we propose and investigate a deep reinforcement learning approach for the optimization of processing paths. The goal is to find optimal processing paths in the material structure space that lead to target structures, which have been identified beforehand to yield desired material properties. The contribution completes the desired inversion of the processing-structure-property chain in a flexible and generic way. As the relation between properties and structures is generally nonunique, typically a whole set of goal structures can be identified, that lead to desired properties. Our proposed method optimizes processing paths from a start structure to one of the equivalent goal-structures. The algorithm learns to find near-optimal paths by interacting with the structure-generating process. It is guided by structure descriptors as process state features and a reward signal, which is formulated based on a distance function in the structure space. The model-free reinforcement learning algorithm learns through trial and error while interacting with the process and does not rely on a priori sampled processing data. We instantiate and evaluate the proposed method by optimizing paths of a generic metal forming process to reach near-optimal structures, which are represented by one-point statistics of crystallographic textures.

Safe and compliant control of dynamic systems in interaction with the environment, e.g., in shared workspaces, continues to represent a major challenge. Mismatches in the dynamic model of the robots, numerical singularities, and the intrinsic environmental unpredictability are all contributing factors. Online optimization of impedance controllers has recently shown great promise in addressing this challenge, however, their performance is not sufficiently robust to be deployed in challenging environments. This work proposes a compliant control method for redundant manipulators based on a superimposition of multiple passive task-space controllers in a hierarchy. Our control framework of passive controllers is inherently stable, numerically well-conditioned (as no matrix inversions are required), and computationally inexpensive (as no optimization is used). We leverage and introduce a novel stiffness profile for a recently proposed passive controller with smooth transitions between the divergence and convergence phases making it particularly suitable when multiple passive controllers are combined through superimposition. Our experimental results demonstrate that the proposed method achieves sub-centimeter tracking performance during demanding dynamic tasks with fast-changing references, while remaining safe to interact with and robust to singularities. he proposed framework achieves such results without knowledge of the robot dynamics and thanks to its passivity is intrinsically stable. The data further show that the robot can fully take advantage of the redundancy to maintain the primary task accuracy while compensating for unknown environmental interactions, which is not possible from current frameworks that require accurate contact information.

The rising availability of large volume data has enabled a wide application of statistical Machine Learning (ML) algorithms in the domains of Cyber-Physical Systems (CPS), Internet of Things (IoT) and Smart Building Networks (SBN). This paper proposes a learning-based framework for sequentially applying the data-driven statistical methods to predict indoor temperature and yields an algorithm for controlling building heating system accordingly. This framework consists of a two-stage modelling effort: in the first stage, an univariate time series model (AR) was employed to predict ambient conditions; together with other control variables, they served as the input features for a second stage modelling where an multivariate ML model (XGBoost) was deployed. The models were trained with real world data from building sensor network measurements, and used to predict future temperature trajectories. Experimental results demonstrate the effectiveness of the modelling approach and control algorithm, and reveal the promising potential of the data-driven approach in smart building applications over traditional dynamics-based modelling methods. By making wise use of IoT sensory data and ML algorithms, this work contributes to efficient energy management and sustainability in smart buildings.

Throughout the world, dust storms wreak havoc on many aspects of human life including health, aviation, solar power generation, and agriculture, among others. Given the hazards that this natural phenomena causes, it is imperative that societies are prepared for the onset of these storms to minimize economic loss and save lives. Utilizing the data received from Earth observation satellites, it is possible for atmospheric scientists to detect developing dust storms; however, even for experts, it can be difficult to detect dust storms in satellite images obscured by clouds, smoke, or nighttime conditions. Furthermore, manual detection requires atmospheric scientists to gather together the relevant satellite images, which takes time before a complete analysis can be made. The ability to automatically detect dust is potentially a large boon for the Earth science community.

Elon Musk is hailed as an innovator and disrupter who went from knowing next to nothing about building cars to running the world's most valuable automaker in the space of 16 years. But his record shows he is more of a fast learner who forged alliances with firms that had technology Tesla lacked, hired some of their most talented people, and then powered through the boundaries that limited more risk-averse partners. Now, Musk and his team are preparing to outline new steps in Tesla's drive to become a more self-sufficient company less reliant on suppliers at its "Battery Day" event on Tuesday. Musk has been dropping hints for months that significant advances in technology will be announced as Tesla strives to produce the low-cost, long-lasting batteries that could put its electric cars on a more equal footing with cheaper gasoline vehicles. New battery cell designs, chemistries and manufacturing processes are just some of the developments that would allow Tesla to reduce its reliance on its long-time battery partner, Japan's Panasonic, people familiar with the situation said.