THINK 2009 Presents

Idea Proposal Category Winners


Finalist: PAUL CHANDRICA MASIH DAS

Lawrence, New York
Lawrence Senior High School

Project Title

LB Trough-Assisted Graphene Synthesis

Abstract

Currently, computer processors are made faster by adding additional silicon transistors, which take up more and more space. Graphene, a two-dimensional carbon sheet, is a recently-synthesized material found to be able to transport electrons at speeds thousands of times faster than any current material. Therefore, graphene transistors might increase computer speed and computational ability without increasing processor bulk. However, current methods of synthesis are tedious, expensive, and produce small, fragile samples. A novel method of graphene synthesis using a Langmuir-Blodgett trough to produce graphene sheets supported on a substrate that could be easy, efficient, and relatively cost-effective is proposed.

MIT Trip Testimony

MIT THINK is much more than a competition. It‘s a life-changing experience that allows you to become part of a community composed of the most intelligent, warmhearted people in the world. At MIT, I felt at home; from watching the Superbowl with complete strangers to playing ping-pong in the dorm. MIT has something unique. It has an extremely diverse community of outstanding individuals who, ignoring all differences, come together as one. Being able to see the best research labs and smartest professors in the country only enhanced my MIT experience, one of the best I‘ve ever had.




Semi-finalist: ZIPENG ZHAO

North Potomac, Maryland
Thomas S. Wootton High School

Project Title

LAO Capable Spacecraft Propulsion System Design

Abstract

As an extension to the current research conducted at Cornell University, this project seeks to develop a power-efficient implementation of Lorentz Augmented Orbit capable spacecraft. Instead of establishing a potential using power supplied by the spacecraft to expel electrons for charging, this project takes upon the engineering approach of using an artificially generated and maintained magnetic field to deflect electrons in the ionosphere onto a charge capacitive module, ameliorating the demanding power requirements. Several quantitative parameters limiting the performance of the spacecraft have been derived and included as functions.

Learning Experience

Through this project, I learned to apply knowledge taught in school to practical concepts of engineering. I also learned to take initiatives in networking with people who share my interest in fields of science. I have become more resourceful in seeking opportunities related to frontier scientific research, and more adept at gathering information from library collections and online databases. The rewards of this experience lie not only in the acknowledgment of my contributions in aerospace engineering, but more importantly in the undertaking of an independent project that stems from a passion to invent and improve.




Semi-finalist: NIKHIL ANAND

Milpitas, California
Monta Vista High School

Project Title

Database Compression through FastOp

Abstract

A new algorithm for compressing static databases and improving operations on these compressed databases is presented. Static databases can be formatted into bitmap indices, which have been used extensively in the past because of their efficiency at answering queries on only low cardinal attributes, such as true or false, male or female, yes or no, and other non-unique values. The author presents a compression algorithm which utilizes a bitmap index compression technique to greatly improve the time performance of search queries. The new search scheme is also shown to work well with high cardinal attributes, which include ID numbers, codes, unique values, etc. This algorithm, called FastOp, compresses a bitmap index through a run-length encoding scheme before processing the search query. These compressed regions are divided into 31-bit groups which are then stored as hexadecimal words, which are inherently less demanding on computational units than bytes or bits, greatly increasing search performance. The efficiency of the algorithm was evaluated by performing logical operations on randomly generated bitmaps (generated by the Markov process) consisting of 10,000, 100,000, and 1,000,000 bits and by comparing time-performance with an existing compression technique known as BBC. The linear scaling of the results shows that the compression algorithm is not only efficient pragmatically, but is also theoretically optimal. FastOp (and in turn, the compression algorithm it implements) has implications in a plethora of fields, but is especially relevant to astronomical physics, particle physics, engineering, and computational science where large databases are usually generated to house scientific information.

Learning Experience

I learned several lessons from my project, having overcome all of the obstacles I was presented with. Perhaps the greatest hurdle I faced was the debugging of the program code. I spent nearly four weeks debugging the program and getting it to perform optimally, and at times it felt like I was trying to push down a rug larger than the floor: each time I set one corner correctly, a bump would come up somewhere else. Also, the mathematical analysis of the program proved very difficult, at least in the early stages. Mathematics is usually the best way to represent the theoretical efficiency of a program and setting up equations to describe my program was definitely a challenging experience. Finally and perhaps the most important lesson that I gleaned from this project: that optimization of an algorithm can be a mathematical and programming monster that truly tests one‘s endurance. As Dr. Donald Knuth once said, "If you optimize everything, you will always be unhappy."