Self-Driving Cars will become more Common

In a study for self-driving cars, a convolutional neural network (CNN) was trained

How data was collected in the study

to map raw pixels from a single front-facing camera directly and translate it into steering instructions. This end-to-end approach proved surprisingly powerful as even with minimum training data from humans, the system managed to learn to drive in traffic on local roads with or without lane markings and on highways. It also operates in areas with unclear visual guidance such as in parking lots and on unpaved roads.

In the United States, thousands of people die as a result of car accidents every year - more than 30,000 people died in 2015. Self-driving cars could, hypothetically, reduce that number as software could prove to be less error-prone than humans. Furthermore, self-driving technology could help individuals with disabilities and the elderly, who are unable to drive themselves.

Unfortunately, possible algorithm failures which may put the safety of pedestrians at risk, is what slowing down the progress and manufacturing of such autonomous vehicles. However, companies like Tesla, are making the public realise that self-driven cars are not a gimmick but rather an innovation designed to improve their lives as well as preventing possible driver errors on the road.

References:

Bojarski, M., Del Testa, D., Dworakowski, D., Firner, B., Flepp, B., & Goyal, P. et al. (2016). End to End Learning for Self-Driving Cars. Retrieved 4 June 2020, from https://arxiv.org/abs/1604.07316

Union of Concerned Scientists. (2018, February 21). Self-Driving Cars Explained. Retrieved 4 June 2020, from https://www.ucsusa.org/resources/self-driving-cars-101