Ten years ago, chips implanted in our heads, commercial trips to space, autonomous and flying cars were the main futuristic novelties for the year 2020. Cut to the current times, some of these technologies already exist, they were executed, but not in a way predominant and popular as we expected.
Speaking specifically of self-driving vehicles, many manufacturers have promised to put these innovations on the streets in 2020, but the truth is that this dream still seems distant.
Special autonomous cars on video
The arrival of autonomous cars is linked to other innovations such as connected, intelligent cities and Internet of Things (IoT), which act as a “right arm” helping vehicles that do not require human interference. While the car is driving alone and communicating with other vehicles, you can read a book, access social media, make a call or simply sleep.
After all, if they were supposed to be available now, what happened? O Tecnoblog listened to the main experts on the subject here in Brazil to understand all the current difficulties, which are not few.
How the autonomous car works and the levels of automation
To function efficiently, the autonomous car depends on five fundamental technologies: cameras, sensors (such as LiDAR), GPS, Artificial Intelligence (AI) and a command center – the car’s brain. In practice, all of them must work to generate accurate responses, such as making the car change lanes at the right time, recognizing traffic lights, objects, pedestrians, animals and acting in dangerous situations.
In addition to the five technologies mentioned, they work with five other levels of automation that help to understand the stage at which a car is in the field of autonomous driving. Know each one of them:
Level 0 – Without automation: the vehicle is dependent on the human being, controlled 100% manually;
Level 1 – Driver assistance: it is the simplest grade and can be found in many vehicles today. In it, the driver can count on features such as Adaptive Cruise Control (ACC), or automatic pilot, and use it as an aid in steering;
Level 2 – Partial automation (semi-autonomous): allows the vehicle to control the steering wheel and pedals alone, so the car can travel without human interference for a few kilometers. Still, the driver needs to be aware to act in risky situations. This level is found in Tesla cars (Autopilot), as well as the Volvo XC60;
Level 3 – Conditional automation: it is an extension of level 2. In it, the car can make detections by sensors and make decisions alone. But the driver must be ready to intervene in cases where the vehicle is unable to execute certain commands. According to the Center for Experimentation and Road Safety (Cesvi Brasil), Uber and Google projects are in this phase;
Level 4 – High steering automation: the vehicle can operate alone and is capable of making decisions thanks to Artificial Intelligence, but it can also ask for your help in a few moments. From this level, the driver can sleep, touch his cell phone or read a book, while the machine works 100% without depending on the human being. But there are barriers: the car at this stage may have difficulty working in adverse weather conditions;
Level 5 – Full automation: it can act without human interference on streets, avenues and highways, in addition to making decisions and even correcting possible failures. At this stage, the vehicle does not need a steering wheel and pedals, and the driver, or rather the passenger, can do other activities inside the car.
Why they didn’t arrive in 2020
The automation levels we showed earlier already reveal the reason for the delay. In the past ten years, companies have spent a lot of money, partnered and dedicated time for autonomous driving to reach 2020. Despite this, how can a self-driving car be highly reliable? That is the question that holds back the arrival of this innovation in the world.
According to the WHO (World Health Organization), annually 50 million people are injured on highways around the world and many of these fatalities happen due to human error. In the future, vehicles that gain autonomy will have to reduce that statistic. Yes, it is a very big responsibility for all organizations working on autonomous car projects.
These companies need to do a lot of training with Artificial Intelligence and the cars have to travel thousands of kilometers and make long trips, without presenting any safety problem. The sensors, on the other hand, need to act without errors dealing with all climatic conditions and identify all objects that may be on the road.
In addition, there is another obstacle in the way: legislation. In Brazil, in addition to the traditional bureaucracy, we come across Article 252 of the Brazilian Traffic Code (CTB), which considers the average infraction “to drive the vehicle with only one hand, except when it is necessary to make regulatory arm signals, change the vehicle’s gear, or activate vehicle equipment and accessories ”.
The main barriers for autonomous vehicles today are basically two: the first has to do with road infrastructure, not only in Brazil, but in several countries, vehicles should communicate with the infrastructure both to position themselves better and to interact more intelligently and with the infrastructure. This is something that is already in development in many places, but that does not yet exist in the necessary magnitude.
The second area that remains to be developed (…) is the legislation itself. In most countries, even though there may be technology for the vehicle to drive alone, (…) the driver still needs to be the last responsible for the operation (…) So, these two pillars – infrastructure and legislation – are the main ones that need to be developed for the full use of autonomous vehicles.
João Oliveira, Volvo Car Brasil general director of operations and innovation
The new coronavirus pandemic has slowed investments, as many manufacturers have revised their current plans to the detriment of the global crisis we are experiencing. When COVID-19 knocked on the door, Ford, which invests billions in technology, reported that its autonomous taxi project was postponed to 2022. The launch is scheduled for 2021.
In June 2020, Mercedes-Benz and BMW paused a historic partnership to develop level 4 autonomous driving technologies. More than 1,000 professionals were involved in the project.
Tesla billionaire and CEO Elon Musk promises a tier 5 car by December of this year, but you better not be so excited about his claim, as we haven’t seen anything very concrete in this more advanced stage so far.
Overall, forecasts are not very optimistic. For now, there is no sign that they will arrive in 2020 or 2021.
Almost all automotive manufacturers today are working on the development of autonomous driving. Ford, Volvo, General Motors, Audi and Tesla are some of them, however, we must pay special attention to Waymo, a company that belongs to Alphabet, a holding company that also controls Google.
Why do we need to keep an eye on her? Waymo is a pioneer in autonomous driving and has a chance to lead this race. In 2018, the company’s driverless cars cover 1.2 million kilometers in California (United States). In January this year, they reached the 32.2 million km mark.
The current scenario is very curious, considering that this race is formed by technology companies and the automotive sector. This situation, however, opens an important discussion: who will put autonomous cars on the streets first? Will Uber, Apple, Waymo and other startups stand out?
The answer is yes and no. Both traditional automakers and technology companies have already presented their first prototypes. A very favorable situation of available cash for investment in research and development weighs in favor of technology companies. For traditional automakers, they weigh more than 100 years of accumulated knowledge in automotive engineering.
Putting wheels on the phone may not be as easy as putting the phone in the car. Anyway, competition is always healthy for evolution and the question is not who will do it first, but who will make it a sustainable business.
Jonathan Marxen, Autonomous Systems mentor at SAE Brasil
Why 5G is so important for self-driving cars
The fifth generation mobile networks will be extremely useful in stimulating the proper functioning of these vehicles. This technology can solve one of the main bottlenecks in autonomous cars today: latency. This means greater safety, since the car will be able to respond more quickly to commands, significantly reducing reaction time.
But not only that. We are also talking about a better interpretation of traffic, better communication with other vehicles and interaction even with traffic lights. Definitely, cars will be smarter and less dependent on humans.
In practice, the long-awaited 5G will enhance V2V, vehicle-to-vehicle intercommunication; the V2I, the vehicle and infrastructure interaction, to analyze, for example, traffic conditions. Or rather, autonomous cars will work with V2X – write this code -, which are connected cars that communicate with everything.
According to Paulo Bernardocki, director of solutions and networks at Ericsson, 5G technology will be essential for communication between vehicles and, at the same time, will help to make decisions.
With this technology, the car communicates with other connected things, stop reacting to the environment and starts receiving additional information on how it can anticipate problems. So, imagine: an accident that occurs a few kilometers ahead, this information can arrive in the car before it approaches the accident (…) and already makes a decision to change the route.
And with all the vehicles connected, we can even think of central traffic coordination, so knowing that many cars driving on a given route, that central coordination makes a decision to spread the flow between two different paths and thus avoid congestion. That’s the importance of 5G with the connected car, you get coordination, you can anticipate things that only the autonomous car without connectivity would not be able to.
Paulo Bernardocki, Director of Solutions and Networks at Ericsson
Is Brazil ready for autonomous cars?
Brazil is the least prepared country for autonomous cars, reveals the KPMG 2020 Readiness Index for Autonomous Vehicles survey. Respectively, Singapore, the Netherlands, Norway, the United States and Finland are the five most capable countries to receive this technology. KPMG recalls in the report that Singapore has invested more than US $ 4.3 million in this innovation, allowed evaluations on its highways and works with massive adoption goals.
Although Brazil enthusiastically adopts new technologies, the country conflicts with several areas that drive autonomous driving and connected cars. In addition, the Brazilian government does not encourage or show interest in putting these projects on the streets. Meanwhile, only companies in the private sector embrace autonomous vehicles for use in their structures.
Brazil still has a lesson to do in terms of infrastructure, legislation, technology and public acceptance. The consultancy KPMG has pointed this out annually (…) The good news is that we are already seeing autonomous vehicles, even if abroad (partial automation has already arrived in cars, trucks and buses). The other good news is that full automation will arrive first in places where technology pays off faster and where there are less regulatory barriers, which are with commercial vehicles traveling in private areas.
Jonathan Marxen, Autonomous Systems mentor at SAE Brasil
Although at a more discreet and timid pace, autonomous cars are on the way. Despite countless challenges and improvements, many companies have proven that it is possible. When are we going to see them on the streets? That will only tell time, but it will not be in 2020!