At first, there was the PC. Then the smartphone came out and revolutionized our society, the way we interact, and how we do daily tasks. Now, it seems only fair to consider the driverless car the next big thing which will shape the future.
Granted, the impact of the autonomous vehicle shapes to be much bigger than the previous two digital revolutions. These connected cars power up development and innovation in other areas as well, not just good old fashioned bent metal.
New data services and Artificial Intelligence (AI) will be two of the most exciting features surrounding the driverless car’s future.
Autonomous Vehicles Will Be Computers on Wheels
When fully autonomous vehicles will be ready for mass production, they will basically be digital sensors generating high amounts of data.
All that data is valuable only if we are able to process it, analyze it, and learn from it. For humans to do so, there would be no efficiency whatsoever. Instead, Artificial Intelligence will take the job.
Cars will undergo a revolution in sensors, which will turn them into endless streams of data. To transfer it for analysis, driverless cars will need to communicate with each other, with different infrastructures, and with the cloud.
Once the data is in the cloud, it will be processed by the AI and served back to all traffic participants in the form of street and even city-level view.
It is expected for AI to be standard on self-driving cars by 2025. It will power features like:
- voice and gesture recognition
- driver monitoring
- virtual assistance
- natural language understanding (NLU).
Thus, drivers will be able to speak to their cars and have them respond and even anticipate certain needs and road conditions.
AI is The Key for Driverless Car Mainstream Adoption
Artificial Intelligence will be essential for the advanced driver assistance systems (ADAS). To work, the machine learning system needs to be fed data from:
- camera-based machine vision systems
- radar-based detection units
- driver condition evaluation
- sensor fusion engine control units (ECU).
After processing and learning from the data it receives, the AI will be capable of detecting and recognizing objects and road conditions, predict actions and adapt.
However, there is much work to be done. We’re just starting to scratch the surface.
Deploying driverless vehicles on public roads requires millimeter precision of surrounding objects, lane information, and direction of movement. There are three options for accomplishing this goal.
For starters, car manufacturers could deploy a fleet of human-driven cars packed with all possible sensors. It is extremely expensive, time-consuming, and will require a commitment towards ongoing updates so that any new change is recorded.
The second option involves semi-autonomous vehicles to carry out this task. It is expensive as well, especially if you consider the amount of sensors needed and their overall cost.
The third option is to use machine learning. It would require an advanced Artificial Intelligence capable of “observing” non-autonomous cars on the road. For example, if a large number of cars perform abrupt steering changes in the same spot, it means there’s an obstacle there.
The same idea can be applied to determine road conditions (by observing when tires are slipping) and weather conditions (by “seeing” when windshield wipers are on).
Who’s In On The Race?
So far we know of the Chinese search engine giant Baidu. They have put $200 million towards AI and AR development for driverless cars and brought in former Microsoft executive Qi Lu as chief operating officer.
On the other hand, DeepScale, a startup out of Mountain View, California, has raised $3 million in seed funding to help automakers use industry-standard, low-wattage processors for more accurate perception.
Their approach includes sensors, mapping, and control systems (“computer vision”) which enable vehicles to make sense of what’s going on around them in real time.
Then there’s ARM, the U.K.-based semiconductor design firm. They have introduced a new chip targeted at markets ranging from self-driving cars to artificial intelligence. The new design by ARM is meant for higher-end IoT applications that must handle complex applications, such as Artificial Intelligence software.
The most important takeaway you should get from this information is that, for once, most industry leaders agree: AI is the future and the key which will make driverless cars a mainstream reality.
Granted, it will take a few more decades until seeing self-driving cars on public roads will no longer be out of the ordinary.
What’s your opinion on Artificial Intelligence and the way it will be used to power autonomous vehicles? Let me know in the comments section below.
If you’ve enjoyed this article, go ahead and subscribe to my weekly newsletter on self-driving cars and all things about the mobility of the future.