|
|
||||||
|
Home |
News |
Articles |
Forum
| |||||||
|
|
Windows Embedded Student ChallengE 2006 -- winning projects
2006-06-25 Article Rating:  / 0
Representatives of 30 student development teams from seven countries converged on Microsoft's Redmond, Wash. campus June 23-25, to compete in the third annual Windows Embedded Student ChallengE, which featured the theme, "Preserving, protecting, and enhancing the environment." Five winning teams collectively were awarded $24,000 in cash.
 The teams, which each comprised three or four undergraduate students plus a faculty advisor, were tasked with designing a computer-based system that solves a "real-world problem" corresponding to the competition's annual theme. The projects were required to run on top of Microsoft's Windows CE embedded operating system, and had to reside on the x86-based eBox II platform hardware (shown at right).According to Microsoft, 367 student teams submitted their products for judging. Of those, 30 projects were selected to compete in the finals, and five finalists were chosen as winners. Among the seven countries represented in the group of 30 finalists, China accounted for nearly half, at 14 entries. The U.S. was second, with five teams, followed by India (four), Romania (three), Mexico (two), and Brazil and Australia with one each, Microsoft said. The five winning projects included systems to: prevent forest devastation; reduce the eradication of protected and rare birds in fish farms; conserve electricity through controlled street lamp lighting; identify and classify bird species; and, improve the lot of coal miners. This article profiles the five winning teams and their projects. Additionally, it provides a list of all 30 finalists, with links to each project's online final report. 44 Tech, Politehnica University of Bucharest, Romania
Team members -- Christian Iuliu Pop, Iaona Romelia Bratie, Omar Choudary, Mircea Dan Gheorghe According to 44 Tech, the most significant cause of deforestation is illegal logging, representing 40 percent of the trees cut in Romania and rising to as much as 80 percent in some countries of South East Asia, Latin America, and Africa. The team's project addresses this problem with a network of sensors placed in the forest, which gather information about cutting noises, temperature, humidity, pressure, and carbon monoxide in the area. Data is processed, and the forestry is automatically informed of problems such as fire or illegal logging.  (Click to enlarge) Forest Watcher incorporates a network of up to 200 wireless sensors covering a forest area up to 200 hectares. The sensors communicate with a Central Unit (CU) built on an eBox II device running Windows CE. Forestry officials receive alerts from the CU on Pocket PC PDAs. The CU is also connected to the Internet. This project's final report can be accessed here (Word file download). Erebus, University of Southern Florida
Team members -- Albert Ng, Jimal Ramsamooj, Francisco Blanquicet, Scott Werner The Erebus Intelligent Scarecrow addresses a problem faced by commercial aquaculture operations that are subject to predation by endangered bird species such as Blue Herons and American Egrets. The Erebus team cites an article from Sept. 2001, claiming that Mississippi Valley fish farms suffer predation losses as high as $18 million annually. Existing techniques such as random noise makers are ineffective, and killing the birds is undesirable from an environmental viewpoint -- and illegal in the case of endangered species.  (Click to enlarge) Erebus uses a video camera and image processing techniques to identify and classify intruders. The eBox II discriminates the type of intruder based on characteristics such as color, and responds in an appropriate manner such as playing high-volume sounds, such as gun shots or bird distress calls, or attempting to physically hit the bird with a high-pressure spray of water. Additionally, the scarecrow performs event analysis, to monitor pest intrusions and gauge the system's effectiveness through its system interface. The system can record events to a log, and provides user notification through email or SMS messaging. This project's final report can be accessed here (Word file download). Stars, Xidian University, China
Team members -- Mingming Cheng, Ling Qiu, Wenbo Li, Shaofu Zhang "Just imagine that in one pitch dark night, there is a brilliant light wave along the lonely road and that when one brightens, another dims out. And the very center of which is you!" Starswave aims to substantially reduce the amount of electricity devoted to nighttime street lighting, by intelligently turning high-power LED lights on and off in response to the presence of pedestrians and vehicles.  (Click to enlarge) Starswave uses infrared sensors to detect vehicles, and a "pyroelectric" sensor to detect people. A series of light-on devices (LOD), each controlling one LED light, is connected through CANbus to an Ebox that controls a street full of lights. By connecting the Eboxes to the Internet, information such as sunrise and sunset times can supplement the control algorithm, as can weather information about fog or rain that may require the lights to be on during daylight hours. The sensors can also provide useful traffic information. This project's final report can be accessed here (PDF file download). The Release Candidates, Politehnica University of Bucharest, Romania
Team members -- Alin Lazar, Andrei Gheorghe, Mihai Ciureanu, Radu Nedelcut The Danube Delta, located in Eastern Romania, is on the list of the UNESCO World Heritage Sites and Biosphere Reserves. It is also home to over 300 different bird species, some of which are very rare or unique, such as the cormorant and the white pelican. BirdSpot is an automated tool for studying these birds without too much "anthropic interference."  (Click to enlarge) The BirdSpot system automatically detects, identifies, and classifies bird species based on visual information, in order to determine the evolution of the population density of distinct species in a designated area. Energy-efficient and affordable wireless devices are placed in remote natural habitats to gather visual data. The data is transmitted to a processing server for aggregate interpretation, based on intelligent image processing and adaptive machine learning algorithms. The output results, consisting of the identified bird species and their locations, are integrated into a database and are accessible via a user-friendly web interface. The team suggests that BirdSpot can also be useful in managing the spread of avian flu, by identifying the arrival of species suspected of carrying the various and automatically notifying authorities. This project's final report can be accessed here (PDF file download). BUPTUNITED, Beijing University of Posts and Telecommunications, China
Team members -- Xingrui Ji, Yi Shi, Lei Wang, Chenpeng Hu Approximately seven million Chinese mineworkers collectively produced over two billion tons of coal in 2005 under very "unstable and hazardous" conditions, according to team BUPTUNITED. Common hazards include explosions, frequent and prolonged exposure to airborne contaminants such as rock dust, and excessive noise and heat stress.  (Click to enlarge) ACES consists of ad hoc, multi-hop, collaborative wireless sensor networks (WSNs) coupled to eBoxes, and ultimately to a central PC that performs complete analysis and permanent storage. The system employs a "three-layered time-critical prediction" strategy that includes sensor layer alarming, CE layer analysis, and finally, PC layer prediction. In this manner, algorithms with different complexities are allocated to different layers, balancing the tradeoff between hardware capability and system requirements. To address the unique requirements of underground mines, the team developed SHARP (threSHold Adaptive Routing by ordering Protocol). SHARP aims to deal with the "belt-like" nature of underground mines and the mobility issues related to workers carrying sensors. This project's final report can be accessed here (PDF file download). Click each project title below, to download the team's final "final report" (PDF or Word file):
|
news feed
