Main Project : A Lie Detector
In week 12, our teacher introduced us to our main project. He informed us that we now have a basic understanding of Arduino and encouraged us to apply our knowledge creatively to develop something useful. This project will contribute to our final grading, and upon completion, we will be required to create a video demonstrating its functionality. Additionally, he mentioned that the most innovative projects will be showcased at a science fair, where numerous attendees, particularly school students, will have the opportunity to interact with our products.
Following the instructions, we formed groups of three individuals. I, Chowdhury Shakline, along with my friends Neelushan Antony and Mahid Ahmed, constituted one such group. After a week of brainstorming various ideas, we collectively decided to create a lie detector.
Initially, we deliberated amongst ourselves regarding the necessary kits, the wiring setup with Arduino, and the coding process. Subsequently, we presented our idea to our teacher, who provided us with the requested kits to facilitate our project development.
At the heart of our project lay the integration of two advanced sensors with Arduino: the AD8232 ECG sensor and a GSR (Galvanic Skin Response) sensor. The objective was to gauge physiological metrics that commonly fluctuate when an individual experiences stress, as in the case of deception. The ECG sensor tracked variations in heart rate, whereas the GSR sensor assessed the skin's electrical conductivity, which typically rises with heightened sweat production induced by stress.
One of the primary obstacles we faced involved the GSR sensor, which initially produced inconsistent and unreliable readings. After extensive troubleshooting, we determined that the sensor was defective. This setback prompted us to acquire a new sensor, underscoring the lesson that sometimes, the remedy is as straightforward as replacing malfunctioning components. Another notable challenge was the calibration of the device to accommodate individual physiological variations. Establishing a universal threshold for detecting deception proved intricate, given the significant variability in stress indicators among individuals. Nonetheless, through iterative testing and numerous trials, we identified threshold values that, while not foolproof, provided a reasonable foundation for our experiments.
Subsequently, we encountered the challenge of presenting the results to the user in a user-friendly manner. While we could view the sensor values on our Serial Monitor, this method was not ideal for the user. To simplify and enhance the user experience, we opted to use two LEDs for output: one green and one red, to indicate truth and falsehood, respectively.
It was essential for our team to recognize the inherent limitations of our project. While the idea of a lie detector is enticing, it's widely acknowledged that such devices are not foolproof. Variables like nervousness or the capacity of certain individuals to remain exceptionally composed can distort the outcomes, resulting in erroneous positives or negatives. This is precisely why lie detectors are no longer admissible as conclusive evidence in courtrooms. Our project served as a hands-on exploration of these concepts, rather than the creation of a definitive truth-determining apparatus. The apex of our project was its exhibition at the STEM fair, where it attracted significant attention. Attendees were enthusiastic about testing the machine, furnishing us with a diverse array of data to analyze. The interactive aspect of the project not only rendered our booth a popular destination at the fair but also instigated lively discussions concerning the science of lie detection, the ethical considerations involved, and the potential for future technological advancements.
Comments
Post a Comment