Project Ideas
The main theme of the RET research project is the development of state-of-the-art methods for wearable systems.
Activity Recognition Using a Wrist Accelerometer (Mentor: Dr. Youn)
Wearable sensor-based physical activity assessments are widely used to evaluate health benefits of physical activities in daily life, and placing an accelerometer to the wrist has become a modern physical activity monitoring research trend in order to improve wear-time compliance for an extended period of time [Mannini, Intille, Rosenberger, Sabatini and Haskell 2013].
RET Research Components
Since the wrist motion is much more complex than the trunk movement, detecting appropriate wrist movements from a single wrist-worn accelerometer is challenging but crucial in developing an accurate physical activity classifier. Our RET teachers will analyze 3-dimensional acceleration data from a single wrist-worn accelerometer to detect a set of physical wrist motions. Teachers will also develop and implement algorithms to classify activities of daily life such as lying, sedentary activities, walking, and running.Corresponding course material to be developed through this research
The ability to analyze, visualize and draw conclusions from large data sets is critical to computing. This unit will allow teachers to learn the process of data collection and analysis in real-world contexts. While differentiating static activities (i.e. sitting or lying down) from dynamic activities (i.e., walking or running), one easy-to-understand classification metric is the standard deviation of vertical accelerations [Youn, Khazanchi, Youn and Siu 2016]. If the standard deviation of vertical acceleration is constantly changing, the subject is performing a physical activity of some type, otherwise he or she is at rest (lying down, sitting, or standing). Based on this observation, a PBL lesson can be developed in the context of a middle or high school math module on statistics. Students learn computing techniques in order to compute mean, median, mode, and standard deviations, which are needed to answer their inquiry-driven questions about the data.
Software Engineering Principles for Wearable Application Development (Mentor: Dr. Siy)
This research topic will focus on techniques for developing reliable applications for wearables. With the increasing popularity of programmable wearables, there is a need for systematic approaches to develop the software needed for such devices [Wasserman 2010]. Designing such systems must account for their limited power. Furthermore, wearables with critical applications, e.g., health monitors, that rely on analog inputs through sensors must be tested thoroughly to avoid failures due to incorrect interpretation of readings.
RET Research Components
We will pursue two threads of investigation: 1) power-aware design techniques and 2) testing sensor-based software applications. Teachers will investigate power-aware software design approaches that quantify tradeoffs between power consumption and time and space efficiency [Stan and Skadron 2003, Jejurikar and Gupta 2006]. Teachers will also investigate techniques for testing applications that rely on sensor inputs to verify if the applications will behave predictably in every possible situation and determine the limits for reliable operation of the wearable application.
Corresponding course material to be developed through this research
Several instruction modules can be developed to enhance the teaching effectiveness of introductory programming courses and impart the value of well-designed and tested programs. Teachers can introduce an active learning module on software design where students investigate power consumption tradeoffs in programming wearables. Teachers can also introduce a module on testing where students learn the value of systematic testing to ensure that applications that serve a critical purpose can be relied upon.
Pattern Extractor on Wearables (Mentor: Dr. Zhu)
This research topic will focus on the exploration and experimentation of a computational scheme on wearables within learning and training environments, emphasizing the valuable patterns embedded in and extractable from the massive amount of data on wearables [Holland 2016].
RET Research Components
This research topic will involve algorithm design and program development (coding) for data collection and pattern extraction by selecting and applying a suitable machine learning technique. For example, smart graphics user interfaces to collect, store, and organize the daily, weekly and monthly data from wearables, clustering techniques to extract the patterns of personal physical activities or social interactions from the data, principal axis factoring approach to analyze and compare the patterns with respect to health or social wellness, and give beneficial advice to the individual. Teachers will learn and acquaint with computational functions on wearable devices, understand and practice with program development cycles from the hands-on research process.
Corresponding course material to be developed through this research
This research project will lead to the development of a teaching module that can be embedded in science or physical exercise courses, or used in a standalone introductory computer science course. By exploring a problem with pattern analysis on wearable devices, students can be guided by the teacher to learn and practice computational thinking in the course module through stages of collecting and analyzing the data, using abstractions and models, and creating computational artifacts.
Wearables in Fashion for Introductory Computing (Mentor: Dr. Morrison)
This research focuses on the use of various wearable technologies in fashion for the purpose of introducing computing concepts to students. Teachers will explore various existing curricular materials for introducing computing through fashion such as those described in [Qiu, Buechley, Baafi, and Dubow, 2013 and Lau, Ngai, Chang, and Cheung, 2009]. They will then develop their own fashion project using available tools.
RET Research Components
This research will involve replicating existing research and designing new active-research components to be implemented in their own classroom. This project will expose teachers to experiment replication and active-research methods to allow exploration within their own classroom. Additionally, teachers will learn information about electrical circuitry and programming involving sensors and continuous polling and interrupts.
Corresponding course material to be developed through this research
This research project will lead to the development of a teaching module that can be embedded in a middle or high school science course (electricity), or used in a standalone introductory computer science course. Exiting curriculum examples will be adapted by the participants for use within their own classroom and for local context. In addition, new curriculum modules will be developed and shared with other participants and in computing education repositories.
Designing Android interface for the answer set solver Clingo (Mentor: Dr. Lierler)
Declarative programming is an excellent vehicle for bringing basic level of understanding of programming and computational problem solving to non-expert users, and answer set programming is a prominent representative of declarative programming approaches that roots in logic programming [Lierler 2016]. Teachers will be trained in this paradigm, tasked to design Android interface to answer set solver Clingo [Gebser, Kaminski, Kaufmann, Ostrowski, Schaub and Thiele 2010], and asked to develop a suitable set of lectures for high school students to illustrate the paradigm in practice.
RET Research Components
Training teachers on basic concepts in answer set programming will bring this declarative programming to a broader cross-section of population allowing early exposure of future professionals to various computational solutions. Teachers will be exposed to several sample domains and then tasked to design a benchmark of their choice, for instance, stemming from the class scheduling domain within their school and model a prototypical solution of this problem in answer set programming. Teachers will also be tasked to design the Android interface to answer set solver Clingo.
Corresponding course material to be developed through this research
Teachers will be asked to develop a set of introductory lectures on answer set programming including solutions to several benchmark combinatorial search problems (e.g. travelling salesman and graph coloring problem).
References
[Mannini, Intille, Rosenberger, Sabatini and Haskell 2013] A. Mannini, SS. Intille, M. Rosenberger, AM.
Sabatini, and W. Haskell. Activity recognition using a single accelerometer placed at the wrist or ankle.
Medicine and science in sports and exercise. 2013 Nov;45(11):2193
[Youn, Khazanchi, Youn and Siu 2016] I. Youn, D. Khazanchi, J. Youn, and K. Siu, Multidimensional
Mobility Metric for Continuous Gait Monitoring using a Single Accelerometer, The 2nd International
Conference on Health Informatics and Medical Systems (HIMS’16), Las Vegas, NV, July 25-28, 2016.
[Wasserman 2010] A. I. Wasserman, Software engineering issues for mobile application development.
Proceedings of the FSE/SDP Workshop on Future of Software Engineering Research, 2010.
[Stan and Skadron 2004] M. R. Stan and K. Skadron, Power-Aware Computing. IEEE Computer, Jan
2004.
[Jejurikar and Gupta 2006] R. Jejurikar, and R. Gupta, Energy Aware Task Scheduling with Task
Synchronization for Embedded Real-Time Systems., IEEE Transactions on Computer-Aided Design of
Integrated Circuits and Systems, 25(6), June 2006, pp. 1024-1037.
[Holland 2016] Janet Holland, Wearable Technology and Mobile Innovations for Next-Generation
Education, IGI Global, April 8, 2016 Education.
[Qiu, Buechley, Baafi and Dubow 2013] K. Qiu, L. Buechley, E. Baafi, & W. Dubow, A curriculum for
teaching computer science through computational textiles. In Proceedings of the 12th International
Conference on Interaction Design and Children (pp. 20-27). ACM, June 2013.
[Lierler 2016] Y. Lierler, What is answer set programming to propositional satisfiability, Constraints
Journal, to appear.
[Gebser, Kaminski, Kaufmann, Ostrowski, Schaub and Thiele 2010] M. Gebser, R. Kaminski, B.
Kaufmann, M. Ostrowski, T. Schaub, and S. Thiele. A user’s guide to gringo, clasp, clingo, and iclingo.
Available at http://potassco.sourceforge.net, 2010