Simplifying the autonomous car
Everyone is talking about the use of artificial intelligence, machine learning and deep learning for self-driving cars. Car manufacturers and tech innovators all over the world envision a day when cars will no longer need drivers. Cars will handle everything, including deft maneuvering, detours and traffic checks. Today, however, there are some glaring challenges that car manufacturers face in making autonomous cars a reality. Even most skeptics agree that they are, indeed, the future of transportation. Experts from IEEE have determined that, by 2040, about 75% of all vehicles will be autonomous and that they will be the most viable form of intelligent transportation.
The last few years have seen almost every tech and automobile company unveil beta versions of their own autonomous car models. While some of the test runs have indeed been promising, let us not forget that we are still in the testing phase.
The recent fatal accidents involving Uber and Tesla Model X are a hard reality check, and they prove that we have a long way to go before the fantasy of true autonomous cars turn into a reality. The fact remains that our roads, our technology and our systems are still adjusting to autonomous cars. Until autonomous cars can completely take over, they will have to coexist with human-driven cars for quite some time. Thus, we need to ensure that we create an environment that simplifies and accelerates the process of bringing autonomous cars onto our roads. Keeping all these challenges in mind, can we flip the problem and try to simplify it for everyone?
Source: Mobiliya
Autonomous car projects are fighting several battles at the same time:
- A challenging regulatory process — While the technology behind self-driving cars has made rapid progress, regulators have been lagging far behind. In a bid to make self-driving cars completely safe, there is the threat of a rise of a complex mesh of rules and regulations that can act as deterrents to the industry. The biggest problem lies in the possibility that each state will have different rules and policies for self-driving cars. Automakers would have a tough time complying with the 50 different requirements of 50 states. In case the federal government does step in to create a national standard for testing, crash liability and design requirements, there might still be individual states and cities that will insist on having the final word, as the roads pass through their territories, making them responsible for the insurance and liabilities related to autonomous cars.
- Technological challenges — Technology must be developed to a point of understanding human intuition. Auto manufacturers and tech companies need to bring AI to such a level that it can substitute for the human mind efficiently enough to replace the human driver, but without replicating human mistakes.
- Geographic challenges — The current global road infrastructure is ill-equipped to handle self-driving cars. Inconsistent traffic signs, driving conditions and traffic laws, for example, are making it unlikely that self-driving cars will become a reality in the near future.
- Liability issues — There are several established rules for dealing with human-driven car accidents. These rules have been based on decades of information about driver behavior. When accidents occur, it is usually clear who is at fault and responsible for paying for the damage. However, in the new era of self-driving cars, liabilities are not clear, and many legal questions may arise.
Accelerating the process of self-driving cars
Instead of fighting all these battles at once, taking a step-by-step approach would be ideal in enabling a fully autonomous car world. A quick review of the existing infrastructure would give us clues about how to make changes to our existing road infrastructure in order to accelerate the process of getting autonomous cars onto the road.
Government: A key player
Ever since the idea of self-driving cars came around, the space has been mainly dominated by OEMs, like GM or Daimler; car-sharing services, like Uber or Lyft; and tech companies like Waymo. However, if the government gets involved, the process of making autonomous cars a reality will be fast-tracked. The government has the power to swiftly create dedicated lanes across the nation and create a common charter of guidelines and safety standards for the design and development of autonomous cars. The central government can call for states to come up with uniform policies, which would reduce the uncertainty and fear of overregulation among manufacturers.
A dedicated lane system
We are already used to having dedicated lanes on the roads for carpooling and public transportation. A similar approach for providing dedicated lanes for autonomous driving can help simplify the process of making autonomous cars a reality. In several European cities, there are train tracks for metro trains that run parallel to city roads. Since a perfect fully automatic vehicle is still some years away, administrations must plan for creating similar dedicated “autonomous vehicle lanes” that are separate from the ones on which humans drive the cars, making their interactions minimal and helping to reduce their chances of colliding. Dedicated test areas and zones marked with alert signs should also be put in place so that human-driven cars, bikes and pedestrians can take extra precautions while driving around these areas. Currently, autonomous vehicles are not equipped to predict human driver behavior, which is very impulsive.
Dollar spend
With billions of dollars being spent on developing self-driving technology, it is critical to ensure that the money is channeled toward the features that promise to advance the “safe” and faster deployment of autonomous cars.
It is clear that self-driving cars will not require the types of roads, traffic patterns and parking that we are accustomed to right now. However, making such radical changes can be an immense task and cannot be done all at once. However, by making simple changes in the lane system, can we simplify and accelerate the process? It is certainly worth a thought.
We need to make sure that autonomous cars see the light of the day before flying taxis start taking over.
All IoT Agenda network contributors are responsible for the content and accuracy of their posts. Opinions are of the writers and do not necessarily convey the thoughts of IoT Agenda.