Will Tesla offer level 5 Autonomy/Self-driving as a service ( AasS ) Soon ?
How Tesla got Self driving right
How are other competitors lagging behind
How expensive is lidar
Why are companies like Uber failing in the race to autonomy
How our brain work while we drive
What an autonomous vehicle
Self-driving vehicles are cars or trucks in which human drivers are never required to take control to safely operate the vehicle. Also known as autonomous or “driverless” cars, they combine sensors and software to control, navigate, and drive the vehicle.
Currently, there is no commercially available fully autonomous vehicle but this can be a reality very soon.
Level 0 - No Automation. This describes your everyday car. ...
Level 1 - Driver Assistance. Here we can find your adaptive cruise control and lane-keep assist to help with driving fatigue. ...
Level 2 - Partial Automation. ...
Level 3 - Conditional Automation. ...
Level 4 - High Automation. ...
Level 5 - Full Automation
How our brain works while we drive
The Frontal lobe handles emotional regulation, while the Parietal lobe is responsible for coordinating visual and touch perception. The Occipital lobe handles the visual response, and the Temporal lobe is for auditory
When you're behind the steering wheel, the Frontal lobe acts as your warning apparatus against impending danger as it alerts your reflexes. The Parietal lobe acts as your gauge on how hard you hit the brakes, or how far you steer during an emergency. The Occipital lobe process everything you see around to know where to steer. Lastly, the Temporal lobe gives you alertness based on what you hear.
Also, we learn to navigate complex scenarios with ease as we learn from scenarios the more experienced the driver is the better the driver can handle emergency scenarios.
the same is applied to AI/Ml algorithms that make the decision to drive using different hardware sensors. Imagine the complexity involved in automation.
“If you recognize that self-driving cars are going to prevent car accidents, AI will be responsible for reducing one of the leading causes of death in the world.”
Let's see different ways companies are trying to reach level 5 autonomy
Telsa
Tesla uses computer vision which involves Camera, Ultrasound sensors, and a combination of the neural net for vision, sonar, and radar processing software for automation.
Upsides
low-cost equipment: The hardware costs approx. $ 1000.
Downsides
Data collections and accuracy: This maybe not the best approach as the hardware like cameras, sensors might not paint the accurate picture for the algorithms and Lidar does a better job at getting the granular details more efficiently. So this requires lots and lots of data for Tesla's algorithms to reach the accuracy of Lidar. There are millions of edge cases the neural nets have to learn how to navigate. Even billions of miles of data might not be even close to a needle in the haystack.
Waymo, Uber, Lyft, etc
Other major companies like Waymo use a combination of Lidar, Sensors, and Geofencing for automation.
Upsides
Accuracy and precision in all weather conditions.
Downsides
Geo-Fencing: It requires detailed 3D maps of the area. which involves a lot of effort in building clear cut 3D maps of cities and roads.
Equipment cost : $40,000 - Autonomous hardware .$7,500 - Lidar Sensor with computer vision comes to around $100,000. This is for sure not affordable for a normal consumer. It might take a lot of time for the hardware to get affordable.
Why I think Tesla is winning the autonomy race
Affordable cost
All Tesla's are already equipped with necessary hardware for autonomy and people have to pay approx. $7000 to use the autonomy features in Tesla.
Data is the new oil
Tesla has now 3 billion miles of data this is far more than any competitor out there. the more data Tesla has the precise and accurate the algorithms become. Real-time data helps neural nets to learn from billions of edge cases around the world.
Autonomy/Self-driving as a service
Taking into consideration the competitive edge Tesla has by collecting all the data from already existing vehicles on the road Tesla can offer the same autonomy as a service to other auto manufacturers.
On average there are around 17 Million new automobiles sold in the US what if 1 or 2 of the major automobile manufacturers use autonomy as a service from Tesla that's close to 20% of the automobile's made which is close to 3 million.
Let's assume Tesla charges the auto manufacturers the service for 1200 a year and that's a whopping 3.6 billion in revenue right out of the box and Tesla also keeps on adding more and more data to make its algorithms more precise and safe.
May be this helps us understand why Tesla is seen more as a software company than an auto manufacturer.
Optimized Insurance with lower rates
Tesla is also planning to create an insurance product to offer its customers which can be revolutionary. The Insurance agencies typically try to accurately price a policy looking at the location, age, driving record, gender usage but that's very limited amount of data. I don't think insurance agencies are doing a good job in pricing the best quotes for individuals looking at the attributes mentioned above. Tesla on the other end knows every bit of your driving habits better than you hence Tesla can analyze and price quotes much more accurately than any traditional Insurance company based on millions of bytes of data they collect which can be a game-changer in the industry.
Conclusion
I feel Tesla is heading in the right direction and would be the first company to reach level 5 autonomy very soon. Introducing autonomy as a service can bring in Tesla tons and tons of cash and data.
If you agree with please comment below. If you don't agree also please comment lets have a discussion.
Yes thats true. Maybe ur cars could earn money for you as robo taxis..!
I think the key will be TaaS, Transport as a Service = Uber as a "Netflix". Pay a static monthly quote a forget private cars, the autonomous car will drive for you... to your office, to your house, to the Supermarket.... 😉