Soiling is a growing area of concern for Photovoltaic (PV) installations in desert areas because of the adverse impact of accumulation of dust on efficiency and reliability of the solar system. The process of conversion of light into electricity is directly impacted by soiling which, as a result, lead to PV performance degradation. Hence, regular PV panels cleaning is crucial to maintain the performance efficiency of PV installations in desert weather condition where sandstorms are a recurring event. Traditional manual cleaning of the panels has the advantage of being cheap, however, it can be time consuming and can cause damage on the PV glass surface when harsh brushing is performed. As a result, further deterioration of PV performance is due to happen. This has led to shifting the focus of research towards exploring alternative cleaning methods that are more efficient and less damaging such as mechanical cleaning, and coating method for PV panel surface.
Additive Manufacturing (AM) technologies have been recently adopted to produce functional spare parts in advanced applications. In this type of applications, high dimensional accuracy is essential, where it is expected for the dimensions of manufactured parts to meet the design specifications and tolerances. Different additive manufacturing processes exhibit different dimensional accuracies. We investigated the dimensional accuracy of additively manufactured metal parts produced by Fused Filament Fabrication (FFF). We designed a new testing structure (Test-Part) incorporating multiple features in a single build. Two different metal materials, 17-4 PH stainless steel, and copper were used to manufacture the test parts. The manufactured specimens were scanned using an advanced laser scanning technology to precisely measure the dimensions and capture the geometry’s details.
Automating the decisions required for creating and setting up a deep learning (DL) pipeline has become a key direction in both research and industry to speed up the complex, slow and error-prone process of designing novel DL architectures across many modalities and tasks. Neural architecture search (NAS) has been a growing trend that aims at automating the design of neural networks that are on par or even outperform hand-designed architectures. While the holy grail for evaluating the quality of an DL model in NAS techniques has largely been focused on getting models with the highest possible accuracy, additional metrics become of paramount importance as AI models meet constrained devices (e.g., mobile and AR/VR systems) or emerging specialized hardware accelerators. Such metrics include speed or computation time, power/energy consumption, memory footprint and model size. To address these additional constraints, there is Cambrian explosion of new research directions: designing optimized deep learning hardware such as low-bit quantization and analog accelerators, efficient neural architecture search algorithms for specialized DL accelerators, and hardware (latency, energy) aware neural network architectures search targeted for constrained devices. This talk uncovers the evolving landscape of hardware-aware neural architecture search and its future directions. It also highlights IBM Research suite of techniques towards the design & build of optimized deep learning hardware as part of IBM’s AI hardware Center initiative.
We will discuss challenges and opportunities in making sustainable fuels and chemicals with energy from the Sun and wind. This talk will focus on recent work towards establishing design rules of the electrode-electrolyte interfacial reactivity for rechargeable lithium batteries, and the reaction kinetics for water splitting and carbon dioxide reduction, by tuning surface electronic structures and solvation environments at the electrified interface.
Climate change is one of the most complex issues interconnected with other global challenges, such as food security, water scarcity, biodiversity depletion and environmental degradation. It is a global problem, felt on local scales, that will be around for decades and centuries to come. Adaptation to climate change involves adjusting to actual or expected future climate by reducing vulnerability towards the consequent harmful effects on our society. Traditional approaches to innovation focusing on one aspect of the problem are proven insufficient and it is imperative to face the problem under a holistic framework. An integrated adaptation approach can be achieved by creating an ecosystem for climate change adaptation solutions, based on a three-tier frame: (a) using the System Innovation Approach (SIA) to integrate multi-faceted technological, digital, business, governance, and environmental aspects with social innovation for the development of adaptation pathways to climate change, to meet EU Green Deal targets for specific regions; (b) linking SIA with Climate Innovation Window (CIW) to form innovation packages by matching innovators with end-users/regions for specific regions; (c) fostering the ecosystem sustainability and growth with cross-fertilization and replication across scales, at European level and beyond, using specific business models, exploitation and outreach actions. The proposed three-tier approach is show-cased in nine widely varied regions across Europe (demonstrators), as a proof-of-concept with regards to its applicability, replicability, potential, and efficacy.