Challenges / Materials Identification Based on Measurements of Passively Emitted Electromagnetic Radiation
FIND Technologies Inc. is a Canadian company that owns novel sensor technology for measuring electromagnetic signatures of materials. The sensor is a robust, inexpensive instrument that detects passive electromagnetic emission from all matter. It has biomedical, homeland security, engineering, geological, and other applications.
In order to develop real-time, automatic identification of materials using the sensor, FIND offers $45,000 prize to an individual or group that is able to invent an algorithm that identifies and categorizes 1,500-sample data set of electromagnetic signatures into 3 classes of substances, at an accuracy of 95% or higher. A Canadian government Informatics Group analyzed three sets of data from our sensor and were able to distinguish one chemical from two others with accuracies ranging from about 70 to 90%, with a statistical classification strategy. We have posted one of their reports which provides the most details concerning methods and results (see Task page).
The contest ends no later than November 30, 2011, or as soon as a successful algorithm is submitted.
The award of $40,000 (Canadian dollars) will be paid by FIND Technologies Inc. to the team who as the first one develops an algorithm that meets or exceeds 95% accuracy threshold. The winning team must transfer intellectual property rights to the algorithm to FIND.
Additionally, there are 5 milestone prizes of $1,000 at 50, 60, 70, 80 and 90% accuracy. The first team or participant who achieves a given milestone will receive $1,000 prize, without intellectual property transfer.
The contest is supported by Biomedical Commercialization Canada, an investor and strategic partner of FIND Technologies Inc.
Details of the task can be found on the Task page. See Leaderboard for current standings. If you have questions or remarks, post them on the Forum.
Declaring the Winner
Please read carefully the Eligibility section of the Terms of Participation to learn about the procedure for selecting a prospective winner and conditions that prospective winner must fulfill to become a winner.
About our Technology
* FIND sensor technology is proprietary.
* A Government Informatics Group, on several occasions has subjected signal data acquired from the sensor to a Statistical Classification Strategy.