Project Sponsor: National Science Foundation (NSF), Division of Undergraduate Education, REU Program
Principal Investigator: Sven K. Esche
Undergraduate Research Assistant: Dennis Hromin
Amount: $12,500
Project Duration: July 1, 1999 - December 31, 2000
Project Summary Sub-project 1: The main purpose of the proposed internet-based remote-controlled laboratory approach is to allow the sharing of costly experimental equipment amongst large student populations. The remote control of the experimental setups in this approach alleviates drastically the strain on the schedule of the laboratory facility by extending its accessibility into the off-business hours of the day. Nevertheless, the individual experimental setups have to be designed carefully so as to minimize the time required for each experiment. This optimization is necessary in order to limit the produced output data (measurement and audio/video data) to an amount that can be reliably handled by the available file storage devices and the computer network. The student working on this particular aspect of the laboratory will develop analytical models of the planned first-order and second-order dynamical systems and advise appropriate combinations of the system parameters using design optimization. Furthermore, the student will test numerically and verify experimentally the proposed parameter configurations. By participating in this team project, the student will gain valuable experimental experience and improved skills in developing simple mathematical models of physical systems and applying easily comprehensible numerical solution techniques for these models.
Project Summary Sub-project 2: The proposed internet-based remote-controlled laboratory approach is based on a client-server architecture where the individual experiments are performed in a batch mode. In this batch-mode environment, the queueing of client requests and scheduling of subtasks becomes an important aspect of the system in order to optimize its performance. The student involved in this specific aspect of the project will develop a program for the queueing of the incoming experimental requests by the clients and test several strategies for the efficient scheduling of the subtasks involved in the processing of one batch job. The student's work will also include an initial investigation of the feasibility of combining identical sets of input data into one single batch job in order to reduce the system load. Using the pilot implementation of the proposed laboratory, the student will use simple analytical and numerical methods to predict the system performance (network load, data transmissibility, response time, etc.) based on the duration and complexity of the experiments. Through this project, the student will be exposed to a research team environment that allows for the application and improvement of analytical as well as experimental skills.