This paper describes a human-in-the-loop evaluation of a power plant Human-Machine Interface (HMI), that was designed using a User-Centered Design approach. General Electric (GE) developed ActivePoint* HMI (AP HMI) to increase plant operator efficiency and awareness. UCD was applied as a design process along with existing industry design guidelines and other design approaches. The goal of the evaluation study was to provide GE Power quantitative data on proposed benefits from the new HMI based on systematic design approaches compared to the legacy HMI product, developed by engineers with an ad hoc process. Compared to this Legacy HMI, the new AP HMI provides enhanced information architecture, simplified graphics and navigation, as well as improved alarm management allowing operators and plant maintenance personnel to focus on what is important. Nine operations staff from a power plant in France performed simulated tasks using both Legacy HMI as a baseline and AP HMI. We measured their performance and satisfaction on the use of HMI to quantify the enhanced effectiveness of AP HMI. AP HMI resulted in 33% higher success rate, 79% lower navigation time, 42% lower total time on task compared to Legacy HMI. Also, the user satisfaction for AP HMI was rated significantly higher than Legacy HMI.

Natural Language Processing (NLP) systems commonly leverage bag-of-words co-occurrence techniques to capture semantic and syntactic word relationships. The resulting word-level distributed representations often ignore morphological information, though character-level embeddings have proven valuable to NLP tasks. We propose a new neural language model incorporating both word order and character order in its embedding. The model produces several vector spaces with meaningful substructure, as evidenced by its performance of 85.8% on a recent word-analogy task, exceeding best published syntactic word-analogy scores by a 58% error margin. Furthermore, the model includes several parallel training methods, most notably allowing a skip-gram network with 160 billion parameters to be trained overnight on 3 multi-core CPUs, 14x larger than the previous largest neural network.

The discovery of novel nanomaterials, such as nanoparticles and nanofibers, is crucial to the expansion of the nanotechnology field. Of even greater importance, is the identification of nanomaterials that exist in nature and have low environmental toxicity when compared to man-made nanomaterials. In 2008, our group first discovered that ivy secretes nanoparticles for surface affixing. It was further demonstrated that these nanoparticles could be used for biomedical applications. This paper proposes an automated framework for high throughput scalable green nanomanufacturing of these naturally occurring nanoparticles. Several parameters necessary to optimize the growth of the ivy, including temperature, humidity, and light level, were regulated using feedback controls. Since the contact of ivy rootlets with a substrate is necessary to initiate the secretion of ivy adhesive, an electromechanical system was designed to automatically stimulate the rootlets to start the nanoparticle secretion process. The contact of ivy rootlets with a surface was formulated as a linear viscoelastic model and a speed control law was proposed for the actuator of the automated system. The proposed framework was verified through prototype experiments, and demonstrated promise for high throughput production of ivy nanoparticles.