Our PresentationReal Self-Driving Car Sensors
Lidar sensors uses a focused beam to measuring millions of points(or more) a second. This sensor also identifies static environment as well as moving objects. It can also detect the ground and lanes.
Camera sensors are normal cameras that rely on light, which are used to spot traffic signals and signs. They are more useful during the day as opposed to night because it can miss things. Many modern cars today also use these camera sensors as rear view cameras such as back up cameras and some side mirror cameras. Medium-range radar, is an inexpensive radar technology that can detect velocity and distance. It can also detect blind spots and/or if other cars are getting close to yours. Ultrasound sensors are short range ultrasonic sensors that used to detect close objects like rear end warning. These sensors also help with parking assist. Long-range radar, is a microwave radar that can measure speed, while also being able to detect obstacles and vehicles up to around 200 meters away. Long-range radar also handle emergency braking, and adaptive cruise control. A Brief History of Self-Driving Cars
The General Motors Futurama exhibit in 1939 featured radio controlled electric cars that drove themselves on the streets.
The General Motors and the Radio Corporation of America tested one of these cars in 1958. They took on 400 feet of a Nebraska highway and embedded it with detector circuits to test the idea, but it didn’t function as they desired. Lidar was first mounted on a car in 2005 for the DARPA urban challenge(driverless car challenge), however, it didn’t finish the race. In 2007 6 cars had finished the race, with 5 of them were using Lidar. Velodyne Lidar is the company making lots of headway in the past. Currently, one of the main important things right now is bringing down the cost of lidar. Now a days we have Tesla autopilot, along with several other almost driverless options on the road. However, Tesla’s autopilot isn’t perfect, but it is far more advanced than anyone could imagine when driverless cars were first thought of. Levels of Vehicle Autonomy
Level 0 - No Automation, meaning the driver is fully in control.
Level 1 - Driver Assisted, meaning the car can alert driver to different conditions. Ex. Cruise Control Level 2 - Partial Assistance, meaning the car can do lots of things with the assistance of the driver. For example, Tesla’s Autopilot. Level 3 - Condition Based Assistance, meaning the car can do the most important things within limited conditions, however the driver must be ready to take control. This is also covered by Tesla’s Autopilot. Level 4 - High Automation, meaning the car can do all important things within limited conditions, and can transfer control whenever. For example you would be able to read while ‘driving’. Level 5 - Entirely Automated, meaning the car can do everything in all conditions, driver can't take control. This means the car would be completely driverless so you could essentially sleep in the car. Current Limitations of Self-Driving Cars
With the current Tesla Autopilots, the driver needs to be pay attention still, and must have a hand on the wheel. If you aren’t always paying attention and having your hand on the wheel, there can be potentially fatal consequences (which there have already been). The first death Tesla Autopilot had was in 2016, the car can’t always react to everything yet. Adverse conditions will also need to have special attention as well.
While prices of lithium ion batteries are decreasing, we are also becoming more aware of the idea that we might not have enough Lithium left. To make 44 million batteries(ie 44 million cars) you would need to have 176,000 tons of lithium. Back in 2016, countries produced 59,300 tons of Lithium. However, we do have reserves left, but in due time (even with recycling) this lack of Lithium will lead to lots of competition which would ultimately lead to higher prices. In order to charge all of the batteries, the United States would need to produce 3080 gWh(gigawatt per hour) every day in order to properly charge all of the batteries. This increase would require the United States to recreate the largest nuclear power plant 33 more times and keep it running 24 hours a day. Potential Benefits of Self Driving Cars
With self driving cars, there is an increased safety for both the pedestrians and passengers of the car, especially the closer we reach to level 5 autonomy. Bellow I will also discuss some more points of how self driving cars will be and can be more safer.
With self driving cars there would be fewer traffic violations, due to the cars programing. Self driving cars will follow all the rules of the road, and wouldn’t violate the speed limit, wouldn’t be diving under the influence as the car would be driving and they can’t get intoxicated, distracted driving/using your cell phones, and violating the rules of the road such as running a red light or reckless driving. These are the most common traffic violations, and they wouldn’t occur with self driving cars as they wouldn’t be able to disobey their coding or rules which would be the rules of the road. Being able to multitask while on long commutes is every persons dream. I definitely know of many people on the east coast who travel hundreds of miles to commute into New York city every single day, and it would take them roughly 2 to 3 hours some times even more. How would you like it if you can get some extra sleep while on your way to work, or if you had a big meeting you have more time to prepare. You would be able to do other things without needing to drive or pay attention to the road IF cars were at level 5 autonomy. Cities would be less congested as cars at level 5 would all be connected (which we will be talking about later), and would be able to communicate with each other on when they will be turning and in what direction. You also wouldn’t have cars blocking the box or intersection as you aren’t allowed to do that. Less/no human-error related accidents as the car would be the one in control and driving, that is if cars are at level 5 autonomy. With self driving cars it removes all human errors that cause accidents such as but not limited to: speeding, distracted driving (i.e. on your phones), driving under the influence, violating the rules of the road (i.e. running a red light, reckless or aggressive driving), and even drowsiness which can occur from long lack of sleep, lock hours and what have you. Self driving cars would reduce the pollution, as they wouldn’t be burring fossil fuels which are harmful to the environment. This is impart due to self driving cars being electric, and due to them running on electricity they wouldn't need to run on fossil fuels like gas. Some Interesting Possibilities with Self Driving Cars
K-City is a city in South Korea that is a fake town that is used for testing self-driving cars.
Connected Autonomous Vehicles, liked I had brought up before. These self-driving cars will not only communicate with each other but also through an interface system. This would mean that all the cars on the road at any given time communicate with one another through a interface system and can tell how fast each car is going, where they are, if they are breaking or will break, have their turn signal on ect. Think of something similar to Uber with absolutely no talking, Meaning self driving cars could be ordered by an app, similar to Uber right now, and the car will be sent to your location to pick you up, and drop you off to your intended location. Automated Cars Economic Impact
Image 19
All of these economic assumptions were made with the idea that 44 million Connected Autonomous Vehicles would take over by 2030. Insurance could lose 60% because they make lots of money from car related issues Trucking could gain 17% because they could convoy easier Automotive could gain 7% because people would still need maintenance and customization Technological Firms could gain 13% because they would be writing the code and making the advancements for it Auto Repair could lose 26% because there would be less accidents Oil & Gas could gain 5% because we would need more electricity to charge all of the electric cars Medical could lose 1% because people would be getting injured less by cars Parking & Road construction could lose 4% because potentially less people would be driving from place to place. Traffic violations could lose 50% because people wouldn’t be breaking as many violations. Legal could lose 1% because there would be less lawsuits related to car accidents. Ethical Acceptance
Self-safety is one of the biggest concerns about self-driving cars as it is about to crash into five people does it favor its two passengers or the five people? These are some of the ethical dilemmas someone would have to answer and code into the car so it could make the right decision. A little later we will talk about the protect me always ethical problems.
One way you can answer this situation is with utilitarianism, which means it is an ethical decision that will also take the greatest goods or in this case the greats number. This means that utilitarianism will always choose the greats number of people to save by determining right from wrong by focusing on the specific outcome or situation it is. Another major question that is asked yet, needs an answer which is when there is a fatal accident or any accident that involves self-driving cars is who to blame? Will it be the company that created the self driving car, the programmer that wrote the software for the car or someone else? Consumer Expectations on Self Driving Cars
Protect me always clause is another big ethical desition/debate as people would want the self-driving car to protect them first. This means if they were the passenger the car would save then no matter how many other people are in the other vehicle. Also if they they were a pedestrian that would be hit, they would want the car to save them rather than saving the people in the car that could have more in it. This would directly contradict the utilitarianism that we talked about earlier. Most people as expect self driving cars to be at level 5 autonomy already, but realistically we are only at levels 2 and 3. People want or expect self driving cars to be at level 5 autonomy right now as they want to be able to sleep in their car while its driving or do other things, rather than actually driving. Another thing is that the older the person the less likely they are to accept self driving cars, thus people would have to convince the elderly accept self driving cars. |
