Isolating single primary rat hippocampal neurons and astrocytes on ultra-thin patterned parylene-C/silicon dioxide substrates

We report here the patterning of primary rat neurons and astrocytes from the postnatal hippocampus on ultra-thin parylene-C deposited on a silicon dioxide substrate, following observations of neuronal, astrocytic and nuclear coverage on strips of different lengths, widths and thicknesses. Neuronal and glial growth was characterized 'on', 'adjacent to' and 'away from' the parylene strips. In addition, the article reports how the same material combination can be used to isolate single cells along thin tracks of parylene-C. This is demonstrated with a series of high magnification images of the experimental observations for varying parylene strip widths and thicknesses. Thus, the findings demonstrate the possibility to culture cells on ultra-thin layers of parylene-C and localize single cells on thin strips. Such work is of interest and significance to the Neuroengineering and Multi-Electrode Array (MEA) communities, as it provides an alternative insulating material in the fabrication of embedded micro-electrodes, which can be used to facilitate single cell stimulation and recording in capacitive coupling mode. © 2010 Elsevier Ltd.

Rigorous finite-time guarantees for optimization on continuous domains by simulated annealing

Simulated annealing is a popular method for approaching the solution of a global optimization problem. Existing results on its performance apply to discrete combinatorial optimization where the optimization variables can assume only a finite set of possible values. We introduce a new general formulation of simulated annealing which allows one to guarantee finite-time performance in the optimization of functions of continuous variables. The results hold universally for any optimization problem on a bounded domain and establish a connection between simulated annealing and up-to-date theory of convergence of Markov chain Monte Carlo methods on continuous domains. This work is inspired by the concept of finite-time learning with known accuracy and confidence developed in statistical learning theory.

Simulated Annealing: Rigorous finite-time guarantees for optimization on continuous domains

Simulated annealing is a popular method for approaching the solution of a global optimization problem. Existing results on its performance apply to discrete combinatorial optimization where the optimization variables can assume only a finite set of possible values. We introduce a new general formulation of simulated annealing which allows one to guarantee finite-time performance in the optimization of functions of continuous variables. The results hold universally for any optimization problem on a bounded domain and establish a connection between simulated annealing and up-to-date theory of convergence of Markov chain Monte Carlo methods on continuous domains. This work is inspired by the concept of finite-time learning with known accuracy and confidence developed in statistical learning theory.

Assessing the accessibility of everyday products

If a product is being designed to be genuinely inclusive, then the designers need to be able to assess the level of exclusion of the product that they are working on and to identify possible areas of improvement. To be of practical use, the assessments need to be quick, consistent and repeatable. The aim of this workshop is to invite attendees to participate in the evaluation of a number of everyday objects using an assessment technique being considered by the workshop organisers. The objectives of the workshop include evaluating the effectiveness of the assessment method, evaluating the accessibility of the products being assessed and to suggest revisions to the assessment scales being used. The assessment technique is to be based on the ONS capability measures [1]. This source recognises fourteen capability scales of which seven are particularly pertinent to product evaluation, namely: motion, dexterity, reach and stretch, vision, hearing, communication, and intellectual functioning. Each of these scales ranges from 0 (fully able) through 1 (minimal impairment) to 10 (severe impairment). The attendees will be asked to rate the products on these scales. Clearly the assessed accessibility of the product depends on the assumptions made about the context of use. The attendees will be asked to clearly note the assumptions that they are making about the context in which the product is being assessed. For instance, with a hot water bottle, assumptions have to be made about the availability of hot water and these can affect the overall accessibility rating. The workshop organisers will not specify the context of use as the aim is to identify how assessors would use the assessment method in the real world. The objects being assessed will include items such as remote controls, pill bottles, food packaging, hot water bottles and mobile telephones. the attendees will be encouraged to assess two or more products in detail. Helpers will be on hand to assist and observe the assessments. The assessments will be collated and compared and feedback about the assessment method sought from the attendees. Drawing on a preliminary review of the assessment results, initial conclusions will be presented at the end of the workshop. More detailed analyses will be made available in subsequent proceedings. It is intended that the workshop will provide workshop attendees with an opportunity to perform hands-on assessment of a number everyday products and identify features which are inclusive and those that are not. It is also intended to encourage an appreciation of the capabilities to be considered when evaluating accessibility.

Natural actor and belief critic: Reinforcement algorithm for learning parameters of dialogue systems modelled as POMDPs

This article presents a novel algorithm for learning parameters in statistical dialogue systems which are modeled as Partially Observable Markov Decision Processes (POMDPs). The three main components of a POMDP dialogue manager are a dialogue model representing dialogue state information; a policy that selects the system's responses based on the inferred state; and a reward function that specifies the desired behavior of the system. Ideally both the model parameters and the policy would be designed to maximize the cumulative reward. However, while there are many techniques available for learning the optimal policy, no good ways of learning the optimal model parameters that scale to real-world dialogue systems have been found yet. The presented algorithm, called the Natural Actor and Belief Critic (NABC), is a policy gradient method that offers a solution to this problem. Based on observed rewards, the algorithm estimates the natural gradient of the expected cumulative reward. The resulting gradient is then used to adapt both the prior distribution of the dialogue model parameters and the policy parameters. In addition, the article presents a variant of the NABC algorithm, called the Natural Belief Critic (NBC), which assumes that the policy is fixed and only the model parameters need to be estimated. The algorithms are evaluated on a spoken dialogue system in the tourist information domain. The experiments show that model parameters estimated to maximize the expected cumulative reward result in significantly improved performance compared to the baseline hand-crafted model parameters. The algorithms are also compared to optimization techniques using plain gradients and state-of-the-art random search algorithms. In all cases, the algorithms based on the natural gradient work significantly better. © 2011 ACM.

Sparsity-Inducing Optimization Algorithm for the Extraction of Planar Structures in Noisy Point-Cloud Data

Most of the manual labor needed to create the geometric building information model (BIM) of an existing facility is spent converting raw point cloud data (PCD) to a BIM description. Automating this process would drastically reduce the modeling cost. Surface extraction from PCD is a fundamental step in this process. Compact modeling of redundant points in PCD as a set of planes leads to smaller file size and fast interactive visualization on cheap hardware. Traditional approaches for smooth surface reconstruction do not explicitly model the sparse scene structure or significantly exploit the redundancy. This paper proposes a method based on sparsity-inducing optimization to address the planar surface extraction problem. Through sparse optimization, points in PCD are segmented according to their embedded linear subspaces. Within each segmented part, plane models can be estimated. Experimental results on a typical noisy PCD demonstrate the effectiveness of the algorithm.

Shock amplification, curve veering and the role of damping

The circumstances are investigated under which high peak acceleration can occur in the internal parts of a system when subjected to impulsive driving on the outside. Previous work using a coupled beam model has highlighted the importance of veering pairs of modes. Such a veering pair can be approximated by a lumped system with two degrees of freedom. The worst case of acceleration amplification is shown to occur when the two oscillators are tuned to the same frequency, and for this case closed-form expressions are derived to show the parameter dependence of the acceleration ratio on the mass ratio and coupling strength. Sensitivity analysis of the eigenvalues and eigenvectors indicates that mass ratio is the most sensitive parameter for altering the veering behaviour in an undamped system. Non-proportional damping is also shown to have a strong influence on the veering behaviour. The study gives design guidelines to allow permissible acceleration levels to be achieved by the choice of the effective mass and damping of the indirectly driven subsystem relative to the directly driven subsystem. © 2013 Elsevier Ltd.

Robotics education: A case study in soft-bodied locomotion

Locomotion has been one of the frequently used case studies in hands-on curricula in robotics education. Students are usually instructed to construct their own wheeled or legged robots from modular robot kits. In the development process of a robot students tend to emphasize on the programming part and consequently, neglect the design of the robot's body. However, the morphology of a robot (i.e. its body shape and material properties) plays an important role especially in dynamic tasks such as locomotion. In this paper we introduce a case study of a tutorial on soft-robotics where students were encouraged to focus solely on the morphology of a robot to achieve stable and fast locomotion. The students should experience the influence material properties exert on the performance of a robot and consequently, extract design principles. This tutorial was held in the context of the 2012 Summer School on Soft Robotics at ETH Zurich, which was one of the world's first courses specialized in the emerging field. We describe the tutorial set-up, the used hardware and software, the students assessment criteria as well as the results. Based on the high creativity and diversity of the robots built by the students, we conclude that the concept of this tutorial has great potentials for both education and research. © 2013 IEEE.