Food Allergen detection with 92% accuracy using BOSCH's BME688 VOC sensor: Is this possible?

For decades, scientists have been searching for a fast and reliable test to identify food allergens. These allergens, typically proteins or carbohydrates, are not volatile by nature, making them difficult to detect quickly. However, a young scientist from the JIC Top 30 has made a groundbreaking discovery. By using the Bosch BME688 sensor, equipped with AI-powered software, he has demonstrated the ability to identify different food allergens with an impressive 92% accuracy. The volatile organic compound (VOC) profile of nuts is particularly complex, as it changes during ripening and roasting (https://www.mdpi.com/2304-8158/10/7/1611). Bosch's BME688 sensor was originally optimized to detect specific gases related to air quality or materials with high volatile organic content (such as coffee). However, the way this young innovator, a Lemelson Foundation award winner ($10,000) and Society for Science Top 30 honouree, leveraged the sensor's AI to detect minute VOCs in nuts is nothing short of miraculous.

Do you think this is achievable, or is the kid made up stuff to win the competition, given that there is no data verification or live demonstration during science fair? I am interested in developing a similar system for detecting allergens, as it could be a game changer for individuals with food allergies.

Watch the video and take a look at the poster to see how accurately the system detects various whole nuts, which contain very low levels of volatile compounds. It appears that several clips were edited together during the testing, which raises some doubts about its authenticity.


https://sspcdn.blob.core.windows.net/files/Documents/SEP/JIC/2024/Posters/2024_JIC_Mahadevan_Samvith_Poster.pdf

https://www.youtube.com/watch?v=TI7SsY7AnEA

3 replies