From the Archives
Archived discussions on CrazyEngineers

A team of researchers led by Ashutosh Saxena, Asst. Prof. Computer Science at Cornell University, have developed a new system that can anticipate the driver's actions so that he or she could be warned against doing it. For instance, the car safety system can predict if the driver is about to take a turn or change lanes and determine if that action could be dangerous. Depending on the system's algorithm that calculates the probability of something bad happening, it warns the driver to avoid it. This means, even though we are some years away from having self-driving cars, this kind of a car safety system can definitely help drivers in having a safe ride. In some of the existing safety systems, the car’s movement is monitored to warn the driver. However, such warning often come too late i.e the driver has already acted. Therefore, the new system developed by Cornell researchers works by observations of driver's bodily movements while keeping the outside environment in context and suggest a proper course of action for the driver.

The new car safety system achieves this by employing cameras to track the driver's body language as well as locating the other vehicles & obstacles on the road. The warning will be issued in the form of sound alert, light alert or vibration (possibly on the left side of steering wheel). Moreover, by utilising GPS systems, the system can also guide the driver if he is about to take a turn towards a one-way street. Smart.


The system has been successfully tested to result in correction prediction of driver’s actions 77.4% of the time ~3.53 seconds in advance. Even though that's the case, such a car safety mechanism has a long way to go before it can be implemented in mainstream four-wheelers. The current system can be easily deceived into wrong predictions. For instance, change in lighting conditions or drivers interacting with passengers or even shadows of passing poles & trees can confuse the system.

To refine the results, the Cornell team hopes to employ various modifications such as infrared cameras for observation at night, tactile sensors for monitoring pressure on the steering wheel, pressure sensors for anticipating braking and 3-D cameras for greater accuracy.

The team is right now all set to present their research work at a workshop on “Model Learning for Human-Robot Communication” at the 2015 Robotic Science and Systems conference slated for July 16 in Rome.

What are your thoughts about having such a car safety system installed on your car? Share with us in comments below.

Source: Cornell Chronicle Research