Intelligent Infrastructure Is Needed For A Driverless Future
A supporting infrastructure for self-driving cars include intelligent structures.
You may hear the media reporting on self-driving cars with emphasis on the car’s technology. This often turns to discussions on AI, LiDAR and computer visioning systems. That is surely an important part of autonomous driving, but there is another component. There must also be an infrastructure in place to support a driverless car.
It is proving to be a challenge for automakers to achieve L5 Full Self Driving certification, let alone get to L3. Honda’s Legend takes the honor of first L3 certified car, but it is in limited release. Even though millions of miles have been logged by self-driving car companies like Waymo, their cars are still not fully capable for safe autonomous driving (e.g. no human driver required).
A supporting infrastructure for self-driving cars include intelligent structures. An implementation of V2X protocols and IEEE 802.11p communications systems aims to increase road safety. This provides a system of control for autonomous vehicles, allowing them to safely navigate roads, bridges, streets and highways using real time data being provided by sensors.
The sensors are installed around roadways and other infrastructure. This gives them access to communicate with self-driving cars and guide them. What this hopefully aims to accomplish is:
- Traffic Efficiency
- Road Safety
If you have been on an amusement park theme ride, you may have noticed that it is highly automated. It is also controlled for safety purposes. The ride takes the visitors to a maze of adventure that is easily navigable. There are no accidents for the most part and it meets safety standards. Applying that concept to intelligent infrastructure for self-driving cars is the next step.
Guided traffic can eliminate accidents when implemented successfully. Just like planes use coordination for takeoff and landing, a network can be developed for self-driving cars using a standard navigation and guidance system. There is complexity in implementing this, and it can come at a great cost. It can be included in infrastructure project for the improvement of roads and highways.
When a car knows its surrounding, it should be able to navigate without problems. If we hypothesize that accident a is equal to 0 in our system:
a = 0
The likelihood of disaster should also be 0. This is because self-driving cars will avoid an accident for its operation, as it will not be a part of its normal routine. That is the way the system has to be trained. It will then rely on guidance from external sensors that provide data for the feedback loop.
The missing piece of the puzzle, for self-driving cars, is the road they are traveling on. Integrating sensors to create intelligent infrastructure can help accomplish fully autonomous L5 driving. Nonetheless, as the AI software continue to improve and cars continue to learn from the accumulated data, the level of confidence for L5 increases. It is just a matter of implementation and acceptance of the system that awaits.
Involved in blockchain development and imaging technology.