While Spielberg’s “Back To The Future” predicted flying cars, what it didn’t anticipate was cars that drive themselves (you’d have to go to “Total Recall” for that one). While purely autonomous vehicles haven’t yet quite hit the market, the idea of having a car that practically drives itself is already a reality.
While you might not think of it this way, your car may have elements in it that are considered the beginnings of automation. Things like cruise control, lane assist, and automatic anti-lock brakes are the beginnings that are evolving into automobiles that propel themselves.
Companies like Tesla, Uber and Waymo are at the forefront of new technology that will soon allow drivers to just be passengers in their own vehicles. However, there are still some kinks to be worked out of the system before cars can be entirely autonomous.
How Self-Driving cars work
In order to make sure that they are safely on the road, self-driving cars use a number of clever technological strategies. The idea is for these to be redundant so as to combine into a relatively foolproof system. Here are the ways that technology is making cars safer:
- Radar: Which can detect objects and their speeds
- Lidar: Using laser reflection data to get better depth perception
- High-Resolution Cameras: These train computer models to recognize objects
- Audio sensors: Assessing depth and proximity through sound
- Networking: Using data to learn decision-making as well as connecting to a larger network of automobiles to learn to map as well as about road and weather.
What challenges still exist?
However, there are a few stumbling blocks yet to master for self-driving cars to really be safe on the road. Here are challenges which still exist with self-driving cars:
Predicting the action of others: Software is not great at envisioning the unpredictable. The idea that other cars or pedestrians will not follow predictable patterns and have human error is confusing to the machine.
Perception complexity: When the software loses sight of something, it has trouble predicting object behaviour. This occurs when objects go into a blindspot, are blocked from view, or get lost in weather like fog or snow.
Cybersecurity: A self-driving car being hacked is a nightmare scenario, but despite all safety precautions, it is probable that a determined expert could hack into these cars.
Are you are looking for more information on what it will take to give the next generation a smooth, driverless, ride? Take a look at the animation below for more details and clarity on how the autonomous car is currently faring.