A man uses machine learning AI, flour, and very loud sirens to fight against package thieves
Youtuber Ryder Calm Down built himself his own anti-package pilfer alarm system using AI, flour, and a truck horn.
The setup is straightforward: a camera monitors the front porch in real time, and a custom model built on TensorFlow (a machine learning platform) detects whether a package is there. Then it goes into self-defense mode. If an unauthorized individual, such as a box robber, removes the package, an alarm sounds, the sprinkling mechanism is engaged, and a blast of flour is shot in the approximate direction of the offender.
This month-long effort entailed teaching the AI to distinguish between a variety of different-sized parcels so that it wouldn't mistake any of them for, example, a stray cat. Ryder also taught the AI to recognize his face as a 'known individual,' thus disarming the alarm system with a whitelist.
Ryder is no stranger to innovative Raspberry Pi projects, which serves as a good reminder of how much more a Pi can do than run emulators. He's already created a dog detector that sounds an alarm anytime a dog walks in front of your house, as well as a secret door that opens with a piano, as shown in the Batman films.
The system, according to Ryder, is designed to deter any would-be terrace thief by creating a lot of commotion.
You can watch the video at this link.