Energy time-series analysis describes the process of analyzing past energy observations and possibly external factors so as to predict the future. Different tasks are involved in the general field of energy time-series analysis and forecasting, with electric load demand forecasting, personalized energy consumption forecasting, as well as renewable energy generation forecasting being among the most common ones. Following the exceptional performance of Deep Learning (DL) in a broad area of vision tasks, DL models have successfully been utilized in time-series forecasting tasks. This paper aims to provide insight into various DL methods geared towards improving the performance in energy time-series forecasting tasks, with special emphasis in Greek Energy Market, and equip the reader with the necessary knowledge to apply these methods in practice.

Serial crystallography at X-ray free electron laser (XFEL) and synchrotron facilities has experienced tremendous progress in recent times enabling novel scientific investigations into macromolecular structures and molecular processes. However, these experiments generate a significant amount of data posing computational challenges in data reduction and real-time feedback. Bragg peak finding algorithm is used to identify useful images and also provide real-time feedback about hit-rate and resolution. Shot-to-shot intensity fluctuations and strong background scattering from buffer solution, injection nozzle and other shielding materials make this a time-consuming optimization problem. Here, we present PeakNet, an autonomous Bragg peak finder that utilizes deep neural networks. The development of this system 1) eliminates the need for manual algorithm parameter tuning, 2) reduces false-positive peaks by adjusting to shot-to-shot variations in strong background scattering in real-time, 3) eliminates the laborious task of manually creating bad pixel masks and the need to store these masks per event since these can be regenerated on demand. PeakNet also exhibits exceptional runtime efficiency, processing a 1920-by-1920 pixel image around 90 ms on an NVIDIA 1080 Ti GPU, with the potential for further enhancements through parallelized analysis or GPU stream processing. PeakNet is well-suited for expert-level real-time serial crystallography data analysis at high data rates.

Deep learning techniques have become one of the main propellers for solving engineering problems effectively and efficiently. For instance, Predictive Maintenance methods have been used to improve predictions of when maintenance is needed on different machines and operative contexts. However, deep learning methods are not without limitations, as these models are normally trained on a fixed distribution that only reflects the current state of the problem. Due to internal or external factors, the state of the problem can change, and the performance decreases due to the lack of generalization and adaptation. Contrary to this stationary training set, real-world applications change their environments constantly, creating the need to constantly adapt the model to evolving scenarios. To aid in this endeavor, Continual Learning methods propose ways to constantly adapt prediction models and incorporate new knowledge after deployment. Despite the advantages of these techniques, there are still challenges to applying them to real-world problems. In this work, we present a brief introduction to predictive maintenance, non-stationary environments, and continual learning, together with an extensive review of the current state of applying continual learning in real-world applications and specifically in predictive maintenance. We then discuss the current challenges of both predictive maintenance and continual learning, proposing future directions at the intersection of both areas. Finally, we propose a novel way to create benchmarks that favor the application of continuous learning methods in more realistic environments, giving specific examples of predictive maintenance.

Many deep learning applications benefit from using large models with billions of parameters. Training these models is notoriously expensive due to the need for specialized HPC clusters. In this work, we consider alternative setups for training large models: using cheap "preemptible" instances or pooling existing resources from multiple regions. We analyze the performance of existing model-parallel algorithms in these conditions and find configurations where training larger models becomes less communication-intensive. Based on these findings, we propose SWARM parallelism, a model-parallel training algorithm designed for poorly connected, heterogeneous and unreliable devices. SWARM creates temporary randomized pipelines between nodes that are rebalanced in case of failure. We empirically validate our findings and compare SWARM parallelism with existing large-scale training approaches. Finally, we combine our insights with compression strategies to train a large Transformer language model with 1B shared parameters (approximately 13B before sharing) on preemptible T4 GPUs with less than 200Mb/s network.

Insects as pollinators play a crucial role in ecosystem management and world food production. However, insect populations are declining, calling for efficient methods of insect monitoring. Existing methods analyze video or time-lapse images of insects in nature, but the analysis is challenging since insects are small objects in complex and dynamic scenes of natural vegetation. In this work, we provide a dataset of primary honeybees visiting three different plant species during two months of the summer period. The dataset consists of 107,387 annotated time-lapse images from multiple cameras, including 9,423 annotated insects. We present a method pipeline for detecting insects in time-lapse RGB images. The pipeline consists of a two-step process. Firstly, the time-lapse RGB images are preprocessed to enhance insects in the images. This Motion-Informed-Enhancement technique uses motion and colors to enhance insects in images. Secondly, the enhanced images are subsequently fed into a Convolutional Neural network (CNN) object detector. The method improves the deep learning object detectors You Only Look Once (YOLO) and Faster Region-based CNN (Faster R-CNN). Using Motion-Informed-Enhancement, the YOLO-detector improves the average micro F1-score from 0.49 to 0.71, and the Faster R-CNN-detector improves the average micro F1-score from 0.32 to 0.56 on the dataset. Our dataset and proposed method provide a step forward to automate the time-lapse camera monitoring of flying insects. The dataset is published on: https://vision.eng.au.dk/mie/

Explainable artificial intelligence (XAI) provides explanations for not interpretable machine learning (ML) models. While many technical approaches exist, there is a lack of validation of these techniques on real-world datasets. In this work, we present a use-case of XAI: an ML model which is trained to estimate electrification rates based on mobile phone data in Senegal. The data originate from the Data for Development challenge by Orange in 2014/15. We apply two model-agnostic, local explanation techniques and find that while the model can be verified, it is biased with respect to the population density. We conclude our paper by pointing to the two main challenges we encountered during our work: data processing and model design that might be restricted by currently available XAI methods, and the importance of domain knowledge to interpret explanations.

Researchers are increasingly using small, autonomous underwater robots to collect data in the world's oceans. Rescuers spotted debris from the tourist submarine Titan on the ocean floor near the wreck of the Titanic on June 22, 2023, indicating that the vessel suffered a catastrophic failure and the five people aboard were killed. Bringing people to the bottom of the deep ocean is inherently dangerous. At the same time, climate change means collecting data from the world's oceans is more vital than ever. Purdue University mechanical engineer Nina Mahmoudian explains how researchers reduce the risks and costs associated with deep-sea exploration: Send down subs, but keep people on the surface.

In recent years, molecular representation learning has emerged as a key area of focus in various chemical tasks. However, many existing models fail to fully consider the geometric information of molecular structures, resulting in less intuitive representations. Moreover, the widely used message-passing mechanism is limited to provide the interpretation of experimental results from a chemical perspective. To address these challenges, we introduce a novel Transformer-based framework for molecular representation learning, named the Geometry-aware Transformer (GeoT). GeoT learns molecular graph structures through attention-based mechanisms specifically designed to offer reliable interpretability, as well as molecular property prediction. Consequently, GeoT can generate attention maps of interatomic relationships associated with training objectives. In addition, GeoT demonstrates comparable performance to MPNN-based models while achieving reduced computational complexity. Our comprehensive experiments, including an empirical simulation, reveal that GeoT effectively learns the chemical insights into molecular structures, bridging the gap between artificial intelligence and molecular sciences.

We introduce collision-resilient aerial vehicles with icosahedron tensegrity structures, capable of surviving high-speed impacts and resuming operations post-collision. We present a model-based design approach, which guides the selection of the tensegrity components by predicting structural stresses through a dynamics simulation. Furthermore, we develop an autonomous re-orientation controller that facilitates post-collision flight resumption. The controller enables the vehicles to rotate from an arbitrary orientation on the ground for takeoff. With collision resilience and re-orientation ability, the tensegrity aerial vehicles can operate in cluttered environments without complex collision-avoidance strategies. These capabilities are validated by a test of an experimental vehicle operating autonomously in a previously-unknown forest environment.

The electron density is a key parameter to characterize any plasma. Most of the plasma applications and research in the area of low-temperature plasmas (LTPs) are based on the accurate estimations of plasma density and plasma temperature. The conventional methods for electron density measurements offer axial and radial profiles for any given linear LTP device. These methods have major disadvantages of operational range (not very wide), cumbersome instrumentation, and complicated data analysis procedures. The article proposes a Deep Learning (DL) assisted microwave-plasma interaction-based non-invasive strategy, which can be used as a new alternative approach to address some of the challenges associated with existing plasma density measurement techniques. The electric field pattern due to microwave scattering from plasma is utilized to estimate the density profile. The proof of concept is tested for a simulated training data set comprising a low-temperature, unmagnetized, collisional plasma. Different types of symmetric (Gaussian-shaped) and asymmetrical density profiles, in the range $10^{16}-10^{19}$ m$^{-3}$, addressing a range of experimental configurations have been considered in our study. Real-life experimental issues such as the presence of noise and the amount of measured data (dense vs sparse) have been taken into consideration while preparing the synthetic training data-sets. The DL-based technique has the capability to determine the electron density profile within the plasma. The performance of the proposed deep learning-based approach has been evaluated using three metrics- SSIM, RMSLE, and MAPE. The obtained results show promising performance in estimating the 2D radial profile of the density for the given linear plasma device and affirms the potential of the proposed ML-based approach in plasma diagnostics.