Guardian Optical Technologies has unveiled a new technology for detecting what drivers and passengers are doing inside the cabin of a moving automobile — all in the name of safety.

The Tel Aviv startup made the announcement at CES 2019, the big tech trade show in Las Vegas this week. Guardian has created Optical Cabin Control (OCC), which focuses on in-cabin car safety. It can monitor the driver, detecting details such as eyelid closure, head pose, whether the driver’s hands are on the steering wheel, and if there is a smartphone in the driver’s hand. It can detect a sleeping baby and make sure the small passenger isn’t forgotten in the car.

Guardian’s sensors detect every moving object inside the vehicle to ensure safety to all the passengers. The OCC product focuses heavily on the driver to monitor their identity and how alert the driver is by measuring the driver’s body position, all in the name of detecting driver drowsiness or other problems that could potentially cause an accident.

“Safety and security of passengers and fellow drivers on the road is our utmost concern,” said Guardian CEO Gil Dotan, in a statement. “Providing a full suite of products that aim to prevent car accidents and the security of passengers in the vehicle is Guardian’s primary focus. Monitoring the driver’s level of alertness is a natural next step in our product development and we are pleased to roll it out to the automotive industry at CES.”

Guardian’s sensors identify the location and physical dimensions of everyone in the car, distinguishing people from objects. By detecting micro vibrations, the system can register, in some cases, a presence even without a direct line of sight. Real-time, comprehensive “big data” is collected from the motion analysis and 3D input, and is fused with image analysis of the sensor’s video feed, to provide a complete analysis of a vehicle’s driver and passengers.

Just one sensor combined with advanced 2D, 3D, and motion analysis can protect drivers and passengers by constantly scanning and tracking occupants and objects anywhere in the vehicle. These technologies work with a car’s seatbelts and airbags to sound immediate alerts. The system deploys machine learning, including image analysis on the sensor’s video feed, as well as big data analysis.

The startup was founded in 2014. It has won major accelerator competitions sponsored by EcoMotion and Microsoft, and has several patents pending, with other patent applications in progress.