Requirements and goals

Requirements

Here is a list of requirements that will be included in the final prototype:

• A hardware unit consisting of a Rasberry Pi that is equipped with motion and proximity sensors that will detect changes in the surrounding environment
• The hardware unit will be equipped with a moderate resolution camera for image capture
• A web application will serve as the user interface including sensor data, captured images, outcome of the analysis, and a log of previous events
• The image capture will be triggered through a remote interface (web application)
• The hardware unit will have Wi-Fi connectivity
• The image capture system will have two modes: fault detection and facial recognition, set locally on the board
• Image processing and/or ML algorithms will be executed in the cloud
• Should have an option to "add people" or id people that are not in the db

Goals

Here is a list of some "nice to have" improvements that may be included in the final prototype:

• The image capture from the camera may be triggered by the motion/proximity sensors
• The image capture system may be able to switch between capture modes from the web application
• The image capture system may have an "automatic" mode, where it tries to classify both fault/facial detection together
• Fault detection should be adaptable to any environment
• Image processing and/or ML algorithms may be run locally on the hardware, if feasible
• Allow the interface to be extendable to multiple devices in the future
• Different camera options, zoom, pan, tilt, would be preferred.