I used to wake up at 5 am in the morning to prepare my 2 kids for school, prepare breakfast and also ready to take them to school as I left to work. Tiresome right?
When they introduced driverless cars, I thought this would make my life easier and now I have day-dreams where every morning I would take my phone and call Mario, my self-driving chauffeur, and direct it to take my children to school and pick my grandmother on it way back and take her to see the doctor.
It would be amazing, but is it safe?
If there is a traffic jam, how does it keep the children safe inside? How does it know the temperature that should be set, does it ask me?
So many questions…
Thus, this article was put together by Handshake to try to answer some of these.
How AI learn (by trial and error)
Please don’t freak out when I say that AI learns by trial and error, it is only natural to fall before you learn to walk. We have seen several incidents with driverless cars, but every time this happens Mario learns a bit more and becomes more autonomous.
The reason why I talk about Mario is because this was one of the best AI learning session published in 2017: an AI software was learning how to play Mario Bros based on trial and error… and it learned well after the character died a few times.
This is a tight analogy to what AI companies are doing with self-driving cars, trial and error. (It is the same thing that Google does with its crawler robots but the “trial and error” part is validated with the click-through rate. There is more on that on this other article Can AI make SEO more human?)
But think with me for a moment: had the self-driving cars existed when cars were first invented, it’s unlikely many people would have thought that humans driving vehicles was a good idea.
All tests are intended to evolve intelligent transport technologies that are the modern technologies used for basic management to drive a car, such as:
- Autonomous cruise control system
- Electronic Differential System
- GPS Controlled Gear Change
- Parking sensors
Autonomous cruise control system
Cruise control was one of the first extraordinary steps towards the heading toward oneself auto, yet dynamic cruise control takes the framework to the following level. You set your sought speed and let the car do the rest. It uses forward-confronting lasers to “see” movement. When it distinguishes a vehicle in front, it diminishes your velocity. At the point when that auto moves over, it continues the first speed.
Independent cruise control or versatile or radar cruise control is a cruise control framework for street vehicles that consequently conform the vehicle pace to keep up a safe separation from vehicles. It makes no utilization of satellite or roadside bases nor of any agreeable backing from different vehicles. Henceforth control is forced focused on sensor data from ready for just. The expansion of helpful journey control obliges either altered framework as with satellites, roadside guides or portable bases as reflectors or transmitters on the again of different vehicles ahead.
These type frameworks go under numerous diverse exchange names as indicated by the maker. These systems utilize either a radar or laser sensor setup permitting the vehicle to abate when approaching an alternate vehicle ahead and quicken again at the present pace when activity permit. ACC engineering is broadly viewed as a key segment of any future eras of sagacious autos. The effect is similarly on driver well being as on streamlining limit of streets by confirming the separation between vehicles as per the conditions.
Electronic Differential System
In car designing the electronic differential is a type of differential which gives the obliged torque to each one driving wheel and permits distinctive wheel speeds. It is utilized as a part of a spot of the mechanical differential in multi-drive frameworks. At the point when cornering the internal and external wheels turn at diverse rates because the inward wheels portray a more modest turning range. The electronic differential uses the controlling wheel summon sign and the engine rate signs to control the ability to each one wheel with the goal that all wheels are supplied with the torque they require.
GPS Controlled Gear Change
The autonomous car can control the gearing automatically. By using GPS the car monitors driving behavior and the position of the car. When need it change the gear automatically and help the driver.
Switching until driver feel the guard pressing on the auto behind helps the driver press into the littlest spaces, yet won’t awe the neighbors. Sensors now beep and blaze as driver close to an article. Numerous autos now offer these both at the back and front and at times at the side. Utilizing camera, these pictures are seen on the dashboard. The most developed models make an elevated realistic of accurate driver position. They’ll even watch out for walker crossing behind the driver.
Although undoubtedly a very advanced and dexterous species, humans don’t always display their best characteristics when in charge of two-ton vehicles traveling at high velocities and in situations where a momentary lapse in concentration can easily kill or maim themselves or others. Humans introduce uncertainty, estimation, and emotion. The relative safety of car travel on a daily basis lulls them into a false sense of security, leading to an ongoing complacency that prematurely ends up to a million lives each year worldwide.
Before we spend the rest of this article looking at the problems that driverless cars might solve and the challenges they will likely create, I want first to explore how we’ve gotten to this point – how are we even able to talk about self-driving cars? What has changed in the world of technology to make it even plausible, and who is at the forefront of this new era?
In this chapter, I will explain at a high level how driverless cars work, and identify some of the key players behind this technology. It’s more accurate to refer to it as a combination of technologies rather than one – the massive parallel advancement of sensors and computing power are both required in various combinations to begin to make this possible. The developments have been pushed primarily by high tech companies, and their rapid progress has driven the auto industry to wake up. Drawing the attention and investment of both Alphabet and Apple, the world’s two most valuable companies, the depth of the changes now sweeping the transport sector begins to become apparent.
Just as many drivers today don’t know much about the operation of their car’s engine, future car “users” (the replacement term for “drivers”) won’t need to know much about how their robotic chauffeurs work. For those interested in understanding, I’ll provide a brief high-level overview of the technologies involved. Perhaps a slightly better understanding of how it works on some level will aid in “trusting” driverless cars, a key factor in uptake that we’ll discuss in more detail later. I think it’s important to have at least a basic understanding of the technologies involved in driverless cars to enable an informed debate about their arrival on our roads.
How Self-Driving Cars Work
The driverless car means the cars which can drive itself without any help of its passenger and also have all the properties to transport the passenger to the desired destination. It’s often called an autonomous car, self-driving car or robotic car. A robotic car is capable of sensing its environment and changes its control. At present, the driverless is available in the market with the highest speed of 20.1 kilometers per hour.
Independent automobiles sense their environment with the help of such system as a camera, sensor, radar, internet, GPS and computer software. Innovative control systems understand neurological information to recognize appropriate routing routes, as well as challenges and appropriate signs. By meaning, autonomous cars are capable of upgrading their charts based on neurological feedback, enabling the automobiles to keep track of their position even when circumstances change or when they get into unknown surroundings.
Levels of Autonomous Vehicles
National Highway Traffic Safety Administration (NHTSA) of USA gives an official classification system for autonomous vehicles in 2013. According to this classification, autonomous vehicles are classified into four different categories based on users of the automatic system. These classifications are;
Level 0: The cars in these categories have no automatic system. The driver has to control the vehicle completely.
Level 1: It has some individual automatic control system such as cruise control, parallel parking, electronic stability control or automatic braking, etc. In these categories, one automatic system can run at a time.
Level 2: More than one automatic system can be controlled at a time such as flexible vacation management with road focusing. Motorists are accountable for tracking the road and are required to be available for management at all periods, but under certain circumstances can disengage from automobile operation.
Level 3: In this categories car, the driver can fully control the car in a certain condition. The automatic sensor fell the situation and provide a message to the driver to control it suitably.
Level 4: The car can control itself properly at safety and critical condition at any time. This vehicle would control all automatically from start to stop. These kinds of vehicle do not need any interpretation from the driver.
Now, we will learn more details about the driverless car. At first, we will know the history of the driverless car from the beginning to till now. Secondly, we discuss the technology used in an autonomous car, and then we will know how it works, how driverless car control and drive the car. We will also know the current cost of the driverless car and its future cost.
Do Self-Driving Cars Need To Crash Before They Can Be Safe?
Most major automakers are pushing the concept of computer-assisted self-driving cars. This certainly has a very high-tech, Star Wars sound to it. Cars that drive themselves? Wow! Sign me up!
Self-driving cars suffer from the assumption that a driver will always be alert and paying attention, ready to take over at a moment’s notice. That poses some significant risks.
The research concluded that a driverless car mobility-on-demand service could mark the end of many families’ need for a second car. KPMG found that people were much less likely to care about zero to 60 miles per hour, horsepower, or torque if they weren’t driving. But they also learned that most people wanted to be able to use these vehicles in the human-operated mode as well.
When asked, “Which automotive brands do you trust to develop self-driving cars?” Google won hands down, with Nissan scoring a decent second and everyone else far behind. Google is seen as the top brand, even though it has never sold a car and obviously is focused on cars that would have neither a steering wheel nor a brake pedal – though that might be a surprise to the survey respondents. While the Google Car is self-driving, it is more importantly designed to be driverless. And that’s quite a big distinction.
Driverless versus self-driving cars
Building a driverless, autonomous car is much harder than building a self-driving car that simply stays in its lane, doesn’t ram the car in front of it, and occasionally passes other vehicles. If you are willing to assume that a driver will be sitting behind the wheel at all times, focused on the road and ready to take over whenever a problem suddenly appears, the technical challenge is far easier. All self-driving cars make this assumption; Google does not. Most self-driving cars are designed to operate in normal conditions, such as far more predictable highway driving. Google Cars are being “taught” to drive themselves in virtually any environment, including complex downtown streets and construction zones.
But consumer research says customers want to be able to assume control of the vehicle at any time. The automakers are listening. The engineers at Google are not, possibly because they believe that in the next five years, as the safety advantages of driverless cars are demonstrated, attitudes will change. After all, it takes only a few million customers to establish a very profitable business and a solid beachhead in a market.
The Google Car is being developed to be fully autonomous. Having Google’s vehicle autonomously drive on highways was largely mastered by 2012. Google’s focus for some time has been the far greater challenge of urban driving, from understanding the intent of pedestrians and cyclists to threading their way through complex construction sites. No other company is close to fully dealing with these intricacies without a human driver ready to take control.
Unfortunately, as long as you have to sit behind the wheel and concentrate on traffic, self-driving systems are not much more than moderately intelligent and very high-priced cruise controls. What Google seeks is far more valuable: the freedom to pay no attention at all to driving and immerse yourself in other interests. As long as you have to remain focused and ready to take over at any instant, is this self-driving much better than simply driving yourself?
Many researchers think that autonomous car would increase the road safety. Most of the cases (95%) road accidents occur due to human error or mistake. A driverless car can reduce it and men would get more time to work.
Research already found that each year there are about 30,000 people deaths in the US only and because of mechanical failure only 5% people death in a road accident. Every year about 1.2 million people are killed by an automobile. So we can reduce 95% accident by proper driving.
The benefits of an autonomous car could include:
- About 1.2 million life $300 billion damages in the US only can be saved by the driving of the driverless car and reduced risk of accidents.
- An autonomous car can drive at high speed that reduces fuel 20%.
- Up to four times as numerous vehicles could go on existing roadways if all vehicles were mechanized. The Texas Transportation Institute says movement blockage squanders 5.5 billion hours and 2.9 billion gallons of fuel every year.
- It would increase the safety on the highway because of its intelligence.
- Traffic control would become easier and could be congestion decreased.
- Vehicle tenants could invest travel time occupied with different exercises, so the expenses of travel time and blockage are decreased.
- Future won’t need conventional parking and used for parking would be reduced only for using automatic parking and reducing total parking costs.
- It can reduce driver costs paid for the taxi driver and commercial transport system.
- Reduce pollution, increase total efficiency and increase proper utilization of power.
- A thief cannot be able to take this due to its intelligence and self-awareness.
- Smooth ride makes journey tiredness.
- No traffic police need to control driverless car and no need for vehicle insurance.
- The driverless car could drive abandoned to wherever it is needed, for example, to get travelers or to go in for support. This would be particularly important to trucks, taxis, and auto imparting administrations.
- The higher speed limit for a driverless car can reduce travel time.
- People can also work in journey time.
- Reduce traffic jam and increase road capacity because of its fewer safety gaps.