In the world of autonomous vehicles, Pittsburgh and Silicon Valley are bustling hubs of development and testing. But ask those involved in self-driving vehicles when they might actually be seen carrying passengers in every city, the almost universal answer is: Not anytime soon.
The Reason We Won’t Have Autonomous Cars Any Time Soon
No current regular-production vehicle should be driven hands free (though there is plenty of footage on YouTube of Tesla drivers doing just that). Part of the reason is that the current semi-autonomous vehicles receive no information from either infrastructure (traffic lights, railroad gates, etc.) or surrounding vehicles.
Even then, there would be older vehicles not yet retired from service gumming up the works. What impact these cars and trucks would have on an otherwise autonomous-ready fleet is difficult to say, but we can assume that even 15 years from now, pure autonomy will likely be a limited condition.
As of this writing, only 10 states have adopted legislation related to the operation of autonomous vehicles. And in each case, those laws relate to the testing of said vehicles on public roads, not the sale and consumer use of them.
The truth is, no matter how autonomous driving becomes, vehicles will still need to be equipped with a steering wheel, throttle and brake pedals, and full instrumentation. This is because human drivers will still be called upon from time to time to take the controls.
With all of the developments in self-driving cars by companies such as Tesla, Waymo, and others you would think that autonomous vehicles would be on the road by now. However, such vehicles are not becoming mainstream and people are wondering what some of the reasons are. Therefore, we would like to explain what is holding up the adoption of self-driving cars and are their solutions to these issues.
Self-driving cars are equipped with AI-powered cameras to help them navigate through the streets and identify some of the objects, road signs and pedestrians they encounter on the road. However, after a heavy snowstorm the camera cannot see the delineation markings on the roads thus making it almost impossible to drive safely. As of today, researchers have not found a way to solve this problem, which is why a lot of the testing is done in warm-weather climates such as San Diego and Los Angeles.
We can add to this the general antagonism that human drivers have towards self-driving cars. In fact, police in one city in Arizona documented at least 21 cases of people harassing Waymo vehicles with one person actually pointing a gun at it.
We mentioned a little bit earlier the general antagonism that people have towards self-driving cars, but the problem exacerbated when we talk about fatalities. A couple of years ago, there was an Uber crash near Phoenix, AZ involving a self-driving car. A poll by AAA determined that 73% of drivers were reluctant to give up control of their vehicles.
We hope that the information presented gave you more insights about why self-driving vehicles have not gone mainstream. Mindy Support is dedicated towards providing researchers with data annotation assistance to help them overcome the problems mentioned above. As AI advances and new developments are made, you can expect to see self-driving cars on the road sooner than most people think.
The idea is that Tesla keeps improving the system by testing through its FSD beta fleet and collecting data to train its neural nets. With enough improvements, Tesla hopes that its FSD system will eventually become safe enough to use without drivers monitoring it, but there have been doubts about making it happen after several missed timelines.
Finally, several smaller companies, including Zoox, a robo-taxi company; Ike, an autonomous-trucking startup; and Voyage, a self-driving startup; have also passed the torch to companies with bigger budgets.
The state-of-the-art semiautonomous cars of the present are as smart and responsive as their artificial-intelligence (AI) algorithms allow them to be. These deep-learning algorithms are trained on insane amounts of data, which helps them recognize an ever-growing number of traffic situations and act accordingly.
Regardless if the timeline is correct, one thing is clear: Reaching Level 5 autonomy will not be a revolution, but rather an evolution of the next generation of sensors that can compete with human vision and an AI that can match human-driver-like reasoning, paired with more robust safety and security features. Once these obstacles are removed, mass adoption will be only a matter of time.
Given the differences between human and cop, we either have to wait for AI algorithms that exactly replicate the human vision system (which I think is unlikely any time soon), or we can take other pathways to make sure current AI algorithms and hardware can work reliably.
One such pathway is to change roads and infrastructure to accommodate the hardware and software present in cars. For instance, we can embed smart sensors in roads, lane dividers, cars, road signs, bridges, buildings, and objects. This will allow all these objects to identify each other and communicate through radio signals. Computer vision will still play an important role in autonomous driving, but it will be complementary to all the other smart technology that is present in the car and its environment. This is a scenario that is becoming increasingly possible as 5G networks are slowly becoming a reality and the price of smart sensors and internet connectivity decreases.
The only relevant metric is not some imaginary and marketing-ish levels, but who will take the financial and criminal responsibility for accidents and death. Note I make a difference between finance and criminal responsibility. Because one can make a case that some deaths from autonomous driving systems will be judged as criminal neglect and at least involuntary manslaughter. In such cases somebody will have to go to prison, not only pay the big bucks.
However, although many Americans have their eyes on the driverless-car prize, slowing the process to allow others to go first may be the wiser strategy. Here are some reasons society should be taking it slow.
One solution is to outfit cars with bigger batteries. But that adds weight, which reduces efficiency. Switching from a gasoline-powered engine to one powered by electricity is another solution. The latter is far more efficient at converting energy stored to power at the wheels. But some of the raw minerals for electric cars come from the Democratic Republic of Congo, a country long dogged by allegations of child labour. Although minerals can be extracted from the sea, this poses political challenges as countries have long bickered over mining rights on the sea floor.
As Musk explained later in the phone hook up: a fully autonomous car will be used five times more than a regular privately owned passenger car, and that delivers a massive amount of potential revenue and earnings from a single vehicle.
The big question is how good autonomous driving will be, and how soon will it be a reality. Musk says it could be proven this year, although it seems doubtful that many regulators will be allowing it to happen on public roads beyond specific trials any time soon. Validation is one thing, approval is another.
Americans spent an estimated 6.9 billion hours in traffic delays in 2014, cutting into time at work or with family, increasing fuel costs and vehicle emissions. Automated driving systems have the potential to improve efficiency and convenience.
The promise of automated cars is that they could eliminate human-error accidents and potentially enable more efficient use of roadways. That sounds, at first blush, like self-driving cars could also mean traffic reduction and lower commute times.
Automation allows real-time traffic information to help cars anticipate what's ahead, then slow down to avoid disruptions, the way the right flow of sand won't overflow a funnel. It also helps that there is no reaction time in an automated car.
Even if flying cars do take to the skies, don't expect to see a world that looks like the one in the "Blade Runner" films. We won't be abandoning four-wheeled vehicles anytime soon. But self-driving cars are clearly on the way, and there are a number of robotized vehicles on display at CES this year.
Some, like the toaster-shaped concept vehicles that fill the Kia booth, look little like the cars of today. The South Korean carmaker envisions a world in which millions of people have abandoned the idea of owning a personal vehicle, relying instead on services like Uber, Lyft and Google spin-off Waymo, which will field fleets of driverless vehicles that can be summoned at the touch of a smartphone app. By pulling the driver out of the cabin, they promise they'll cost less than having a car in the driveway.
But the Kia concept cars, like those from several competitors, envision massive changes to the interior of tomorrow's vehicle. If you don't have to sit behind a steering wheel, why not turn the cabin into a mobile living room, office, even a hotel room?
But even some proponents are pushing back. Only a few years ago, former Nissan CEO Carlos Ghosn had predicted his company would be selling its first fully hands-free models by 2020. But Denis Le Vot, chief executive of Nissan North America, acknowledged in an interview at CES that the time frame was too optimistic. Part of the reason is technical, much of it is cost-related, the challenge being the need to "make the technology affordable," he said.
CES has become an alternative showplace for automakers that traditionally might have displayed their new cars, concepts and technologies at events like next week's North American International Auto Show in Detroit.
There are indeed still fundamental challenges to the safe introduction of fully autonomous cars, and we have to overcome them before we see these vehicles on our roads. Here are five of the biggest remaining obstacles.
In the future, machines will be able to do this detection and classification more efficiently than a human driver can. But at the moment there is no widely accepted and agreed basis for ensuring that the machine learning algorithms used in the cars are safe. We do not have agreement across the industry, or across standardisation bodies, on how machine learning should be trained, tested or validated. 2ff7e9595c
Comments