- Envision Team
The Future of Autonomous Vehicles
Imagine having a personal chauffeur that drives you everywhere. Imagine decreasing your transportation time by half because you don’t need to wait at stop lights anymore. The future with autonomous vehicles seems promising and optimistic. In the present time, autonomous vehicles have become the next big thing. Companies and VC’s are pouring billions and billions of dollars into this industry. The topic of autonomous vehicles involves both technical and ethical discussions. Currently, is our AI technology advanced and trustworthy enough to implement these robots into society? There has been frequent reporting of traffic accidents due to autonomous vehicles. For the advancement of such a technology, is it worth it to place human lives at stake? This article will briefly focus on the introduction of autonomous vehicles in our society today, how far it has developed, and the companies currently in competition for development.
Why Autonomous Vehicles?
For a more efficient future
No more traffic
This cuts transportation time
As accidents are human errors, by introducing robots and taking humans out of the equation, and assuming that these robots work, there will be fewer accidents
Companies are in competition with each other to see who can build the most practical & functional autonomous vehicle
Billions and billions of dollars are completely invested in this sector
However, it is also super expensive to develop (cars are expensive in general)
It has been argued that more money should be put into the development of AI software
Currently it is 80% mechanical/20% software
More and more new cars have cruise-control as an added mode
It is taking quite a long time to develop: https://www.nytimes.com/2019/07/17/business/self-driving-autonomous-cars.html
Fully autonomous-vehicle-ified communities were predicted to exist by 2021
Currently there are no national nor industrial standards for autonomous vehicles
Transport car service companies (Uber, Lyft, Waymo) are losing money: https://www.theverge.com/2018/8/15/17693834/uber-revenue-loss-earnings-q2-2018
Uber lost $891 million the second quarter of 2018
Engineering Advancements in the Field
Deep learning techniques
Computer vision algorithms
SLAM and other mapping techniques
Identifying objects, roads
LIDAR & other sensors
Auto-pilot/self-parking mode in vehicles
How can we fully integrate autonomous vehicles into the modern world?
Bureaucratic challenges: revolutionizing current traffic, pushback from car companies and other possible competitors
How will this technology deploy across the world? Should it vary by country or should it be unified?
Economic challenges: possible job loss across numerous industries (trucking, freight, rideshare, etc.)
Is the process of development justified?
To test autonomous vehicles, we must put them in the real world, which means real people and real lives at stake. There also might be structural damage.
The cars must be on actual roads with unexpected obstacles in order to “learn” and gather data
Public disapproval: https://www.caranddriver.com/features/a29587219/autonomous-vehicle-enemies/
Technical barriers: requirement for massive amounts of data, low latency decision-making, increased need for cybersecurity
Companies/Startups in the Field (in no order)
BMW has an Autonomous Driving Campus that specifically tests self-driving cars, developed in 2018.
Hyundai has invested 4 billion dollars into self-driving cars.
Amazon has two relevant projects: an autonomous package delivery service, and Amazon Web Services, a software service that permits one to run difficult computations on an Amazon computer.
General Motors plans to expand into San Francisco for self-driving cars.
Uber has raised 1 billion dollars for development in this field.
Lyft is expanding its test sites globally.
Waymo, a spin-off of Google, has cars in American cities for testing and development.
Faraday Future, a California-based company with most operations in China, is reimagining the "typical" automotive experience, from user interface to algorithmic routines.
Tesla is well-known for its autopilot feature, which allows the driver to take control at any time.
A future with autonomous vehicles is not impossible, but there are so many more obstacles than the public actually predicts
In order for the development of this technology to proceed, the vast majority of the nation has to be on board
People will disagree, and that is completely normal, but even if the public is disagreeing with the technology, how can it advance? The autonomous vehicles impacts their daily lives the most
AI is still at its initial stages
In order for AV to work smoothly, they must act like humans and think like humans
Requires the cars to make inferences (top-down thinking)
Cool Links, Projects - Miscellaneous Finds
Hyundai is creating autonomous vehicle Ubers in California. Check out their recent work at this link: https://www.theverge.com/2019/10/25/20932237/hyundai-self-driving-car-ride-hail-irvine-date