Sequencing of multi‑robot behaviors using reinforcement learning

Given a collection of parameterized multi-robot controllers associated with individual behaviors designed for particular tasks, this paper considers the problem of how to sequence and instantiate the behaviors for the purpose of completing a more complex, overarching mission. In addition, uncertainties about the environment or even the mission specifications may require the robots to learn, in a cooperative manner, how best to sequence the behaviors. In this paper, we approach this problem by using reinforcement learning to approximate the solution to the computationally intractable sequencing problem, combined with an online gradient descent approach to selecting the individual behavior parameters, while the transitions among behaviors are triggered automatically when the behaviors have reached a desired performance level relative to a task performance cost. To illustrate the effectiveness of the proposed method, it is implemented on a team of differential-drive robots for solving two different missions, namely, convoy protection and object manipulation.     

Pilot proteomic analysis of cerebrospinal fluid in Alzheimer's disease

Proteomic analysis of cerebrospinal fluid (CSF) holds great promise in understanding the progression of neurodegenerative diseases, including Alzheimer's disease (AD). As one of the primary reservoirs of neuronal biomolecules, CSF provides a window into the biochemical and cellular aspects of the neurological environment. CSF can be drawn from living participants allowing the potential alignment of clinical changes with these biochemical markers. Using cutting-edge mass spectrometry technologies, we perform a streamlined proteomic analysis of CSF. We quantify greater than 700 proteins across 10 pairs of age- and sex-matched participants in approximately one hour of analysis time each. Using the paired participant study structure, we identify a small group of biologically relevant proteins that show substantial changes in abundance between cognitive normal and AD participants, which were then analyzed at the peptide level using parallel reaction monitoring experiments. Our findings suggest the utility of fractionating a single sample and using matching to increase proteomic depth in cerebrospinal fluid, as well as the potential power of an expanded study.     

Characterization of PA-12 specimens fabricated by projection sintering at various sintering parameters

Large area projection sintering (LAPS) promises to be a new method in the field of additive manufacturing. Developed in the Mechanical Engineering Department, University of South Florida, LAPS uses long exposure times over a broad area of powder to fuse into dense, reproducible materials. In contrast, LS, a common powder-based additive manufacturing, uses a focused beam of light scanned quickly over the material. Local regions of concentrated high-energy bursts of light lead to higher peak temperatures and differing cooling dynamics and overall crystallinity. The mechanical properties of laser sintered specimens suffer because of uneven particle fusion. LAPS offers the capacity to fine-tune fusion properties through enhanced thermodynamic control of the heating and cooling profiles for sintering. Further research is required to identify the relationship between LAPS build settings and part properties to enable the fabrication of custom parts with desired properties. This study examines the influence of LAPS sintering parameters on chemical structures, crystallinity, mechanical, and thermal properties of polyamide-12 specimens using powder X-ray diffraction, Fourier transform infrared spectroscopy, differential scanning calorimetry, small-angle X-ray scattering, scanning electron microscopy, and microhardness testing. It was observed that higher crystallinity was imparted to specimens that were sintered for a shorter time and vice versa.     

On matrices, automata, and double counting in constraint programming

Matrix models are ubiquitous for constraint problems. Many such problems have a matrix of variables $\mathcal{M}$ , with the same constraint C defined by a finite-state automaton $\mathcal{A}$ on each row of $\mathcal{M}$ and a global cardinality constraint $\mathit{gcc}$ on each column of $\mathcal{M}$ . We give two methods for deriving, by double counting, necessary conditions on the cardinality variables of the $\mathit{gcc}$ constraints from the automaton $\mathcal{A}$ . The first method yields linear necessary conditions and simple arithmetic constraints. The second method introduces the cardinality automaton, which abstracts the overall behaviour of all the row automata and can be encoded by a set of linear constraints. We also provide a domain consistency filtering algorithm for the conjunction of lexicographic ordering constraints between adjacent rows of $\mathcal{M}$ and (possibly different) automaton constraints on the rows. We evaluate the impact of our methods in terms of runtime and search effort on a large set of nurse rostering problem instances.     

On tasks synchronization with the MMS protocol

The MMS protocol is a powerful tool for communication as well as for synchronization between manufacturing equipment such as robots, programmable logic controllers, etc. This paper shows how the MMS services can be used to handle tasks synchronization in a distributed environment. The MMS weaknesses regarding temporal aspects and possible directions to alleviate these weaknesses are also presented.     

Self-organizing maps for pattern recognition in design of alloys

A combined experimental–computational methodology for accelerated design of AlNiCo-type permanent magnetic alloys is presented with the objective of simultaneously extremizing several magnetic properties. Chemical concentrations of eight alloying elements were initially generated using a quasi-random number generator so as to achieve a uniform distribution in the design variable space. It was followed by manufacture and experimental evaluation of these alloys using an identical thermo-magnetic protocol. These experimental data were used to develop meta-models capable of directly relating the chemical composition with desired macroscopic properties of the alloys. These properties were simultaneously optimized to predict chemical compositions that result in improvement of properties. These data were further utilized to discover various correlations within the experimental dataset by using several concepts of artificial intelligence. In this work, an unsupervised neural network known as self-organizing maps was used to discover various patterns reported in the literature. These maps were also used to screen the composition of the next set of alloys to be manufactured and tested in the next iterative cycle. Several of these Pareto-optimized predictions out-performed the initial batch of alloys. This approach helps significantly reducing the time and the number of alloys needed in the alloy development process.     

Inducing imperfections in germanium nanowires

Nanowires with inhomogeneous heterostructures such as polytypes and periodic twin boundaries are interesting due to their potential use as components for optical,electrical,and thermophysical applications.Additionally,the incorporation of metal impurities in semiconductor nanowires could substantially alter their electronic and optical properties.In this highlight article,we review our recent progress and understanding in the deliberate induction of imperfections,in terms of both twin boundaries and additional impurities in germanium nanowires for new/enhanced functionalities.The role of catalysts and catalyst-nanowire interfaces for the growth of engineered nanowires via a three-phase paradigm is explored.Three-phase bottom-up growth is a feasible way to incorporate and engineer imperfections such as crystal defects and impurities in semiconductor nanowires via catalyst and/or interfacial manipulation."Epitaxial defect transfer"process and catalyst-nanowire interfacial engineering are employed to induce twin defects parallel and perpendicular to the nanowire growth axis.By inducing and manipulating twin boundaries in the metal catalysts,twin formation and density are controlled in Ge nanowires.The formation of Ge polytypes is also observed in nanowires for the growth of highly dense lateral twin boundaries.Additionally,metal impurity in the form of Sn is injected and engineered via third-party metal catalysts resulting in above-equilibrium incorporation of Sn adatoms in Ge nanowires.Sn impurities are precipitated into Ge bi-layers during Ge nanowire growth,where the impurity Sn atoms become trapped with the deposition of successive layers,thus giving an extraordinary Sn content (＞6 at.％) in Ge nanowires.A larger amount of Sn impingement (＞9 at.％) is further encouraged by utilizing the eutectic solubility of Sn in Ge along with impurity trapping.     

Hydrogen detection using platinum coated graphene grown on SiC

The characteristics of hydrogen detection using epitaxial graphene covered with platinum are reported. The multi-layered graphene was grown by chemical vapor deposition (CVD) on a Si-polar 4H-SiC substrate. Surface morphology was characterized by scanning electron and atomic force microscopy. Current-voltage measurements and real-time monitoring of the current flow through the graphene/platinum device were used to confirm the response to hydrogen gas. The background temperature was varied from room temperature to 175 °C in order to measure the activation energy of hydrogen detection.