Autonomous Cars

McDonald’s article advances the idea that Teslas Company could be offering a fully autonomous car by the end of 2020. Elon Musk believes his company, Tesla, will produce a fully autonomous vehicle by 2020. Musk is sure that there are currently no critical challenges to Tesla’s goal to give the world a fully self-driven car. The article highlights five challenges that face the realization of a fully autonomous vehicle. The first problem is developing a sensor that can operate in all weather conditions experienced anywhere on earth and in congested cities like Hanoi and Cairo. Autonomous cars rely on multiple sensors to “see” the environment, detect objects, and road signs. Heavy traffics, lousy weather, and road signs that have graffiti on them can adversely affect these sensors’ accuracy.

Another problem is with machine learning. Autonomous cars rely on machine learning and AI to process data and make decisions. However, there is currently no accepted basis to ensure that machine learning algorithms applied in cars are fully safe across the industry. There is also the need to ensure that the vehicle adjusts effectively to open the road and does not forget previous safety behaviors. The fourth problem comes with regulation; the automobile industry still has no adequate standards and regulations for a fully autonomous car. The existing standards assume that a vehicle has a human driver to take control in an emergency. The last problem pointed out is the social acceptability of the autonomous car. The current tests for Tesla autonomous cars and other self-driven vehicles have recorded high-profile accidents, raising a social acceptability issue for the car buyers and those sharing the road.

Response

Even though Elon Musk, the CEO of Tesla, believes that the company is ready to produce a fully autonomous car, a fully autonomous vehicle that can self-drive safely and legally without human input will take the past 2020, that dream may still not be realized. More improvement and thorough testing are necessary before fully autonomous cars become part of everyday life, traffic, and gain social acceptability. As impressive as Tesla’s trials are, most of them are typically carried out under ideal environments. The current tests are conducted in fenced or extremely controlled settings-fully mapped in 3D. However, we all know that the situation with traffic in the cities is anything but predictable, characterized by “chaos”—pedestrians jumping on and off footways, cyclists creatively maneuvering their ways, and drivers shifting lanes, making turns, or stopping unexpectedly. The current fully autonomous car technologies cannot match these challenges, and it will take a while to be ready for the problems. As such, Musk termed as “many small problems” may not just be small but sophisticated also. However, does it mean that producing a fully driverless car that can complete a legally and safely journey without human intervention is out of range for the anticipatable future? The answer is simply no.

Extensive researchers and automobile companies strive to bring the underlying autonomous car technologies and machine learning to the next level, making them reliable and robust. With the current research and developments, autonomous cars covering all impending conditions and applying cases with 99.999% reliability and precision need more time, maybe not 2020, but in the near future.

References

McDermid, J. (2020). Autonomous cars: five reasons they still aren’t on our roads. https://theconversation.com/autonomous-cars-five-reasons-they-still-arent-on-our-roads143316