Venkateswaran Shekar

Shared Research Resources (SRRs), also known as Core Facilities or Research Service Centers are essential for advancing scientific research in academic, government, industry, and non-profit institutions since they provide access to advanced resources like innovative research technologies, latest equipment in variety of platforms, and expert personnel with skills to solve complex problems as well as providing training and pursue technology development. Traditionally, SRRs are thought of as confined to academic R1 universities and non-profit organizations. However, many regional centers and national laboratories also feature core facilities that cater to local, regional, national, and international users. Like traditional SRRs, these facilities provide user training, in-person access, and remote access to instruments and services. In this section, we will discuss some of the opportunities these cores provide to scientists worldwide. We will also discuss some of the challenges these facilities face in running and maintaining such a large and continuous operation with many users.

Machine learning is a useful tool for accelerating materials discovery, however it is a challenge to develop accurate methods that successfully transfer between domains while also broadening the scope of reaction conditions considered. In this paper, we consider how active- and transfer-learning methods can be used as building blocks for predicting reaction outcomes of metal halide perovskite synthesis. We then introduce a serendipity-based recommendation system that guides these methods to balance novelty and accuracy. The model-agnostic recommendation system is tested across active- and transfer-learning algorithms, using laboratory experiments for training and testing and a time-separated hold out that includes four different chemical systems. The serendipity recommendation system achieves high accuracy while increasing the scope of the synthesis conditions explored.

Autonomous experimentation systems use algorithms and data from prior experiments to select and perform new experiments in order to meet a specified objective. In most experimental chemistry situations, there is a limited set of prior historical data available, and acquiring new data may be expensive and time consuming, which places constraints on machine learning methods. Active learning methods prioritize new experiment selection by using machine learning model uncertainty and predicted outcomes. Meta-learning methods attempt to construct models that can learn quickly with a limited set of data for a new task. In this paper, we applied the model-agnostic meta-learning (MAML) model and the Probabilistic LATent model for Incorporating Priors and Uncertainty in few-Shot learning (PLATIPUS) approach, which extends MAML to active learning, to the problem of halide perovskite growth by inverse temperature crystallization. Using a dataset of 1870 reactions conducted using 19 different organoammonium lead iodide systems, we determined the optimal strategies for incorporating historical data into active and meta-learning models to predict reaction compositions that result in crystals. We then evaluated the best three algorithms (PLATIPUS and active-learning k-nearest neighbor and decision tree algorithms) with four new chemical systems in experimental laboratory tests. With a fixed budget of 20 experiments, PLATIPUS makes superior predictions of reaction outcomes compared to other active-learning algorithms and a random baseline.

Several transportation network vulnerability models have been proposed. However, most only consider disruptions as a static snapshot in time and the impact on total travel time. These approaches cannot consider the time-varying nature of travel demand nor other undesirable outcomes that follow from transportation network disruptions. This paper proposes an algorithmic approach to assess the vulnerability of a transportation network that considers the time-varying demand with an open source dynamic transportation simulation tool. The open source nature of the tool allows us to systematically consider many disruption scenarios and quantitatively compare their relative criticality. This is far more efficient than traditional approaches which would require days or weeks of a transportation engineers time to manually set up, run, and assess these simulations. In addition to travel time, we also collect statistics on additional fuel consumed and the corresponding carbon dioxide emissions. Our approach, thus provides a more systematic approach that is both time-varying and can consider additional negative consequences of disruptions for decision makers to evaluate.

Transportation networks are critical to the social and economic function of nations. Given the continuing increase in the populations of cities throughout the world, the criticality of transportation infrastructure is expected to increase. Thus, it is ever more important to mitigate congestion as well as to assess the impact disruptions would have on individuals who depend on transportation for their work and livelihood. Moreover, several government organizations are responsible for ensuring transportation networks are available despite the constant threat of natural disasters and terrorist activities. Most of the previous transportation network vulnerability research has been performed in the context of static traffic models, many of which are formulated as traditional optimization problems. However, transportation networks are dynamic because their usage varies over time. Thus, more appropriate methods to characterize the vulnerability of transportation networks should consider their dynamic properties. This paper presents a quantitative approach to assess the vulnerability of a transportation network to disruptions with methods from traffic simulation. Our approach can prioritize the critical links over time and is generalizable to the case where both link and node disruptions are of concern. We illustrate the approach through a series of examples. Our results demonstrate that the approach provides quantitative insight into the time varying criticality of links. Such an approach could be used as the objective function of less traditional optimization methods that use simulation and other techniques to evaluate the relative utility of a particular network defense to reduce vulnerability and increase resilience.

The trend towards ever more powerful and faster processors has led to an enormous increase in power consumption. This paper focuses on scheduling tasks in a heterogeneous environment with DVS enabled processors to minimize both execution time and energy consumed. The proposed algorithm, called Energy-Dynamic Level Scheduling (EDLS), favors low-energy consuming processors by introducing a cost factor that affects scheduling decisions. Our scheme allows for trade offs between energy consumption and the desired performance. Our simulation results exhibit significant power savings at a reasonable increase in overall execution time. Moreover, our results demonstrates a high degree of correlation between the energy saving and the increase in the heterogeneity of processors.