ARTIFICIAL INTELLIGENCE, THE BRAIN OF SELF DRIVING CARS (TESLA).

“Once You Trust A Self-Driving Car With Your Life, You Pretty Much Will Trust Artificial Intelligence With Anything.” —Dave Waters”

According to hopkinsmedicine.org, the brain is a complex organ that controls thought, memory, emotion, touch, motor skills, vision, breathing, temperature, hunger and every process that regulates our body. As such in Self Driving Cars its brain is that complex Software System that powers and enhances the efficient operation in terms of movement, environment understanding etc. For Driver-less Vehicles, its Artificial Intelligence.

Self Driving Cars is a possibility because at the forefront of achieving this future of safe and fully autonomous vehicles is TESLA. Additionally, the said company have been rolling out features upon features in an attempt to advance the Frontiers of their Safe and Fully Driver-less Car dream, but there exist LYFT, BMW, FORD, MERCEDES, VOLKSWAGEN alongside in the struggle with all of them equipped with very Talented engineers working to achieve this future.

ARTIFICIAL INTELLIGENCE

What powers the Tesla Car in terms of it being Autonomous is Artificial Intelligence. AI which is short form for Artificial Intelligence is not what you think it is and broad way Media have actually made out headlines to hype and unintentionally scare people about the topic. Artificial Intelligence is simply building machines that can make decisions in the world without it being explicitly programmed. It can also be defined as training computers or machines to learn from experience and the computer in turn using that experience in solving problems it was trained for.
Manually writing a program to make computers distinguish between pedestrians and moving vehicles is a very near impossible task. On the flip side, programming computers to learn from the experience of a driver, which includes stopping by the Zebra Crossing, not hitting any nearby vehicle, ability to identify a pedestrian and a moving vehicle etc teaching a machine from billions of such example which is the experience, rather than manually programming it is what AI is all about. Learning from experience and applying the experience.

HOW DO COMPUTERS SEE THE WORLD

Humans have the ability to look at images or videos and understand it instantly, but for computers it not the case because they are basically “dumb and stupid” boxes.
What a computer understands is numbers and images are represented in the computers brain as a gigantic grid of numbers between [0, 255].
What powers the ability for cars to navigate the road by itself, is a concept in AI called Computer Vision. Computer Vision is all about building machines that can recognize, interpret and make decisions about what has been presented to it. Computer Vision gives eyes to machines, by enabling machines understand the real world like we humans do.

SELF DRIVING CARS

According to Wikipedia.com, Self Driving Cars also known as an autonomous vehicle, driver-less car, or robotic car, is a car incorporating vehicular automation, that is, a ground vehicle that is capable of sensing its environment and moving safely with little or no human input. It uses sensors, computers and robotic actuators to automate some or all driving tasks.

BRIEF HISTORY OF SELF-DRIVING CARS ——2000’s - Till Date

This article appreciates many of mans contribution towards Driver-less Vehicles in the 1900’s but wont consider it for now rather, focus will shift to the the 2000’s.
The years 2004,2005,2007, 2010s, 2020s remain very significant for Self Driving Cars in the following ways.
In 2004, the U.S. Department of Defense’s Research Arm, DARPA organized series of Competitions/Challenge, that was tailored towards getting engineers and researchers or people with interest in Autonomous Driving to build such vehicles that can self-navigate 150 miles of roadways in desert American Countryside. In the end out of 106 teams nobody won the One Million Dollars Prize money.
In 2005, there was a follow-up competition and interestingly, 106 was increased to 195 meaning 89 new competitors joined on. In the end by October of that year, the final competition was held only five teams were able to finish but “STANLEY” Sebastian Thrun’s robot car won the race finishing in about seven hours. What wasn’t possible in 2004 and before became possible and the driver-less vehicle problem was solved.
In 2007, another challenge followed and this time around the Driver-less were required to complete a route in an Urban Environment, while expected to get by traffic jams and handle situations like parking and road intersections and consequently adhering to road traffic ordinance in California.
In conclusion, out of 11 teams selected for the finals, November 2007 (Southern California) only six finished the challenge and the winner after the 240 minutes challenge came from Carnegie Mellon University,.
In the 2010’s, Google also revealed its self-driving cars after public roads tests have been carried on it over an over again on the other side, powerful and generational lasting automobile companies like Ford Motors, BMW, Mercedes-Benz also made attempts to get involved in Self-Driving Technology.
According to reports Uber a rideshare company backed out of the Self Driving Car race due to loss of money, law suits appertaining to patent infringement, misappropriation of trade secrets, that bordered around a former Google engineer that used technical information from Google to build his self-driving startup Otto that was acquired by Uber.
Currently, Tesla Motors is leading this race and its the closest company at the commercial level putting self-driving capabilities in the hands of its customers.

TESLA ?

Tesla is an Automobile company that manufactures Electric powered cars. Founded by Mark, Os, Teducx and Elon Musk the Super Human co-founded the company in the year 1999 and since then the company is now worth 1 Trillion Dollars. According to them, all their cars built since October 2016 come standard with a feature that powers Self-Driving called the AUTOPILOT.

TESLA’S AUTOPILOT?

Autopilot is basically a software system that can be found in every modern Tesla Vehicle. Simply defined, it is a feature with an autonomous driver assistance system that allows a Tesla to recognize the lane its driving by interpreting Cars, Pedestrians interactions etc, while being capable of making decisions its being programmed to perform example: driving, parking, holding brakes, auto parking, smart summon of car etc. This doesn't make a Tesla vehicle fully autonomous self-driving car but rather requires human intervention when needed.

ARTIFICIAL INTELLIGENCE COMPONENT OF TESLA’S VEHICLE

What enables all of this is an Artificial Intelligence algorithm called “HydraNets” enabled by Neural Networks which can performs varying tasks. The Engineering explanation behind TESLA’S AI, will not go deep rather two questions I believe is necessary to understand the workings.

• What makes up the AI Setup of Tesla’s ability to see and drive?
• How does a Tesla have a proper understanding of its environment?

The Engineering explanation behind TESLA’S AI, will not go deep rather two questions I believe is necessary to understand the workings

What makes up the AI Setup of Tesla’s ability to see and drive?

1. raw information processing and representation: According to http://tesla.com a modern Tesla vehicle is equipped with 8 external cameras, 12 ultrasonic sensors and a powerful onboard computer which provides an additional layer of safety to guide the user during commute. The Algorithms bottom layer (raw), processes raw sensor data from the cameras mounted on the vehicles. The information from this 8 cameras is translated into a 3-Dimensional ‘Vector Space’ which can be ascribed to a simulation of the real world which contains road signs, cars, humans etc that feels like a video game.
2. backbone/neck: The outputs from “raw” is received by the “backbone” which is a Computer Vision Algorithm called a Convolutional Neural Network that is responsible for recognizing information within the images its processing and then it outputs very rich high level features information that is then used in conjunction with the “neck” that also serve to extract more greater detailed information.
3. head: This is the upper part of the Network which makes use of information its fed from the previous layers to perform tasks like Lane Prediction, Traffic Lights Task, Object Detection.
4. Tesla’s HydraNets uses a combination of State of The Art AI Algorithms meant for Computer Vision tasks, and Natural Language Processing tasks.

How does a Tesla have a proper understanding of its environment?

1. Really right! yeah, the food of AI is data and it is trained on data in order to be able to make decisions in the world or wherever its used in
2. DATA: Tesla is trained on lots of data
3. Requesting the Fleets
4. Creating More Training Data

CONCLUSION

In the end Artificial Intelligence is the brain of Self-Driving Vehicles and the future cannot be any brighter because with major breakthroughs in Computer Visions Research, coupled with companies like Tesla turning research results to products and putting them at the commercial level consequently, we have nothing else but firm believe in the possibility that Cars that are Safe and Fully Autonomous will soon be the norm in society. This doesn't undermine other companies though but by taking a detailed look at what Tesla has done any Machine Learning Engineer will agree that it’s mind blowing—solving very difficult Artificial Intelligence problems at scale and bringing it to the market.

REFERENCES

1. Mark Harris, “Ethics, Technology, and Public Policy” – Autonomous Vehicles, set of materials developed by an interdisciplinary research team at Stanford University, led by Hilary Cohen, Rob Reich, Mehran Sahami, and Jeremy Weinstein.
2. The Road To Conscious Machines, The Story of AI by Micheal Wooldridge.
3. Two Minute Papers, Watching Tesla’s Self-Driving Cars Learn, youtube.com.
4. Jason Zhang, “Deep Understanding Tesla FSD”, saneryee-studio.medium.com/deep-understanding-tesla-fsd-part-1-hydranet-1b46106d57.
5. Benars Marr, “How Tesla is Using Artificial Intelligence to Create The Autonomous Cars Of The Future”, linkedin.com/pulse/how-tesla-using-artificial-intelligence-create-autonomous-marr.