Search

In this paper we describe an NSF-TUSE funded project, which seeks to expose a broad group of undergraduate students to data science methods via already existing introductory undergraduate STEM classes (as opposed to through the creation of entirely new classes). A key component is a novel process, described here, to develop data science modules that naturally fit into such existing STEM classes. This novel process takes most of the development burden off the instructor's shoulders and instead employs collaborative teams of undergraduate and graduate students as well as the PIs in consultation with the instructor. The outcomes of this process are data science modules ready to be infused throughout the undergraduate STEM curriculum. We describe best practices and lessons learned about the module development and deployment process as well as several of the modules and their fit into the curriculum. In this paper we furthermore describe data science experiential learning opportunities for students at more advanced stages of their education. We finish by outlining the methods by which we will be evaluating the effectiveness of both the modules as well as the experiential learning opportunities in exposing students to data science methods. [ABSTRACT FROM AUTHOR]

We present the Fidget Car - a one degree-of-freedom car driven by a DC motor - for use in small group laboratory exercises or classroom-based small group learning activities of 10-40 minutes in length. These activities can be directed towards a number of STEM courses, ranging from undergraduate mathematics or introductory engineering to systems and controls classes in electrical and mechanical engineering. The motivation for developing these activities is to enable students to develop intuition regarding core systems and controls concepts prior to or integrated with presentation of mathematical concepts and analysis techniques: the activities are designed to enable students to "visualize" the mathematics of systems. We provide an overview of the Fidget Car design, a materials list, example activities and use cases for the Fidget Car, and a pointer to a Google site archiving materials for reproducing and using the Fidget Car. We illustrate approaches for data acquisition and analysis that incorporate only a smartphone, a laptop, and open-source physics software, enabling activities to be conducted in a classroom setting. [ABSTRACT FROM AUTHOR]

Detection of hidden surface crevasses on glaciers is a vital process involved in over-snow traverses for science and resupply missions in Polar regions. There are several areas warranting improvement in the current protocol for crevasse detection, which employs a human-operated ground penetrating radar (GPR) on a mid-weight tracked vehicle. In this fashion, a GPR scout team must plan an appropriate crevasse-free route by investigating paths across the glacier. This paper presents methods supporting a completely autonomous robotic system employing GPR probing of the glacier surface. We tested and evaluated three machine learning algorithms on post-processed Antarctic GPR data, collected by our robot and a Pisten Bully in 2009 and 2010 at McMurdo Station. We achieved 82% classification rate for a linear SVM, compared to 82% using logistic regression and 80% using a Bayes network for contrast. We also discuss independent versus sequential classification of GPR scans, and suggest improvements to or combinations of the most successful training models. Our experiment demonstrates the promise and reliability of real-time object detection with GPR. [ABSTRACT FROM PUBLISHER]

This paper presents methods to continue development of a completely autonomous robotic system employing ground penetrating radar imaging of the glacier sub-surface. We use well established machine learning algorithms and appropriate un-biased processing, particularly those which are also suitable for real-time image analysis and detection. We tested and evaluated three processing schemes in conjunction with a Support Vector Machine (SVM) trained on 15 examples of Antarctic GPR imagery, collected by our robot and a Pisten Bully tractor in 2010 in the shear zone near McMurdo Station. Using a modified cross validation technique, we correctly classified all examples with a radial basis kernel SVM trained and evaluated on down-sampled and texture-mapped GPR images of crevasses, compared to 60% classification rate using raw data. We also test the most successful processing scheme on a larger dataset, comprised of 94 GPR images of crevasse crossings recorded in the same deployment. Our experiments demonstrate the promise and reliability of real-time object detection and classification with robotic GPR imaging surveys. [ABSTRACT FROM PUBLISHER]