Machine Learning at Road Intersections

Computers are ubiquitous to everyday life, and increasingly machine learning is being implemented to solve everyday challenges that improve the human condition where once human decision-making was required.  Yet, the utilization of machine learning is still lagging in many essential aspects of our everyday life. 

Every year pedestrians in my community become victims to inattentive drivers. Drivers education provides drivers with the experience and knowledge in order to avoid many accidents, but sometimes accidents are unexpected. People may not be aware of their surroundings when driving or are currently focused with something else. That is something Drivers education warns against; however inattention to the road still occurs. Five years ago, my friend was run over while crossing a busy intersection coming home from school. Machine learning used effectively can significantly reduce the pedestrian casualty rate. Many intersections are already equipped with speed and red-light cameras to enforce existing traffic laws, but we can add to their capabilities.  We can automatically link every smartphone to these cameras, redirecting phones to an easily downloadable app. Using the intersection cameras, the app can distinguish between static and mobile objects using a program that learns how to act most judiciously from data patterns. The app would make recommendations to the driver about how to respond most effectively if an identified object proves to be in the predicted path of the vehicle. For example, when it senses that a driver intends to disregard a red light, it can encourage the driver to stop. It can also alert the driver that a pedestrian or cyclist is approaching by using biometric machine learning algorithms that can distinguish between moving objects. The intelligent app would improve itself by re-assessing its efficacy and adjust its method of responding in response to each interaction. Through mathematical modeling, vehicle operators would get customized advice, when pertinent, regarding the risk that is associated with failing to comply with the app’s recommendations. The future of machine learning seems boundless; apps that can learn from experience and improve with usage represent a clear future to the incorporation of machine learning in improving our everyday life.

The gruesome site of lifeless bodies lying on the cold hard streets in intersections throughout town and even near my high school is an awful reminder of how quickly life can be lost, unnecessarily. Incorporating technology into driving is essential because it can act like an expert driver ed instructor seated next to the driver to help as a second set of eyes and as an advisor.  This may be most beneficial to new and inexperienced drivers, to tired or distracted drivers but really to anyone who is behind the wheel of a vehicle.  It can even be adapted to self-driving vehicles if and when those become more widely used. The combination of both Drivers Education and technology will create a future in which people can feel equally safe at home and on the road. I hope to one day take part in creating machine learning technologies that can save members of both my community and others. That is why I am majoring in computer science. I want to make a profound difference on the human condition in our country.