Autonomous Cars - Dream to Reality!
W

hen I was 12 years old, I thought that autonomous vehicles were a distant dream. Engineers  would have to scrutinize and study mounds of data. Furthermore, would have a piece of machinery capable of generating an output, that is sophisticated enough to account for the complications of the road. If autonomous vehicles were indeed introduced to society, it’s structure would constantly be modified and altered to make it easier for self-driving cars to navigate.  A few researchers had even stated that they would not create large self-driving vehicles till they’ve made principal changes to our roads. On top of the time it would take to create and curate these cars, a multitude of accidents are bound to occur. Even with an unfathomable number of test runs, there is no guarantee that sensors and AI systems of a car will interpret every moment with accuracy. This could lead to mass car pile ups, and disastrous accidents. As a result, manufacturers would have to implement key protection features from various safety evaluations. This was a fiction I gambled with in the past. Something that you see in futuristic films, but not introduced to the real world. But yet again, revolutionary engineers prove that even the most far fetched ideas can lead to feasible ideas.


Today, we see that autonomous vehicles are not just a reality, but the future of transportation at a global scale. This new technology is being adopted by China, the U.S.A, Canada, and other countries across the world. Uber, Tesla and Waymo's self-driving cars are leading us to a future of safer roads. Researchers have tested results from a multitude of fuel-efficient cars on miles traveled, and found - upon giving the option to drivers - that under many circumstances, people choose to drive cars that had a greater fuel-efficiency. Self-driving vehicles, having higher fuel efficiency (with the ability to take ideal routes to locations), will prove to be able to market itself as a more eco-friendly choice. The United States is largely involved in AI implementation and development. They have taken on a massive role in the development of autonomous vehicles in recent years,  and have also become a role model to other countries when it comes to community readiness. The autonomous vehicles readiness index (AVRI) indicates that the U.S was ranked 4 in 2019. Making sure our roads are ready for autonomous vehicles is nearly as important as the technology put inside them, as roads are the environment we maintain for autonomous vehicles.


There are 2 companies that have taken a front row seat in this technological revelation today. These companies are Tesla and Waymo. As we close in on our  driverless future, people have begun to theorise the future competition and quality projected by these two multinational corporations.  Waymo, a company dedicated to building - the world’s most experienced driver;  and Tesla, a company that produces both the software and hardware involved in the development of an autonomous vehicle - they both provide us with the latest developments and insight into the machinery and technology inside today’s self-driving vehicles. Both of these companies have managed to take self-driving vehicles to the next level, and in this article I will be introducing you to the key features of Waymo and Tesla’s hardware and software.


Elon Musk, CEO of Tesla

2016 was a step in the right direction for Tesla. That year alone proved to be an incredible  contribution to the future of automation. Tesla was able to introduce auto-piloted systems in Tesla’s very own model S, 3 and X cars. A step towards level 5 automation.  With 8 cameras providing a 360 degree view of the vehicle’s surroundings, 12 updated ultrasonic sensors and an improved front circuit radar to decode the environment and landscape (improvement on the previous model) that may be in the vicinity of the car - Tesla has truly demonstrated the complexity of autonomous vehicles. Today, in 2020, AI autopilot is our new take on complete automation. Details released by Tesla introduce 5 key factors that are being integrated into newer models of Tesla’s electric cars. These factors include hardware, neural networks, autonomy algorithms, code foundations and evaluation infrastructure. These 5 features are what make Tesla’s autopilot cars a reality.  A critical piece of hardware included in the AI autopilot systems include a silicon chip, designed to perform and accelerate AI driven calculations. Neural networks function through data created by per-camera networks that analyze the road surrounding a vehicle. With 48 neural networks placed and integrated into a single autonomous vehicle, there is a vast amount of data gathered on every drive. Autonomy algorithms help fuel the data evaluated by the car’s neural networks, “producing high-fidelity representation of the world and planning trajectories in that space”. Code foundations are also built, running the autopilot systems that are integrated into the hardware and software of Tesla’s self-driving vehicles. These codes allow for smooth, clean transactions between systems. And finally, evaluation infrastructure, this software tracks the performance of Tesla vehicles currently in operation. It supports and enhances system progress, and discourages regression. Tesla is taking large steps towards our goal of level 5 automation, producing massive ramifications.


Waymo Self Driving Cars

Waymo has been dedicated to the autonomous vehicle journey since 2009, when the Google self-driving car project began. Waymo, as opposed to Tesla - who use autopilot as the key to automation - use Lidar to automate their cars. Lidar, also called “laser scanning”, is an eye-safe laser that creates 3D representations of a surveyed environment. Typical Lidar sensors emit waves that bounce off objects and return to them, allowing a car to register the distance between itself and the objects in its vicinity. Waymo’s Lidar (Laser Bear Honeycomb) has “unmatched design for 95degree field of view” (FOV). With millions of miles driven on public roads, and billions in simulation, Waymo has been able to test their technology’s capabilities in a multitude of environments. Before releasing a car into an environment however, Waymo’s team constructs a detailed map, pinpointing a road’s features including signs, markers, curbs, sidewalks and lights. The car then uses it’s sensors to detect changing elements in an environment to make sure it can navigate safely around obstacles. Software created by Waymo assumes the correct trajectory, speed and steering required to make that successful journey.


Conclusion

Waymo and Tesla have made extraordinary contributions to the future of vehicle autonomy, and continue to do so through these trying times - by discovering new technologies that will lead to level 5 autonomy. Their goals are the same but their paths are quite different, following two very distinct routes to find the answer to full autonomy. As level 3 cars  hit the markets now in Japan, we can expect greater discoveries to unravel in the coming years. Positive —the close to future

references
  1. Velodyne Lidar. Url
  2. Waymo. Url
  3. Tesla. Url