I'd like to talk about an idea everyone in the self-driving car world knows, but that you may be unfamiliar with... it's called Autonomous Vehicle Platooning.
What is Platooning?
Imagine you own a fleet of trucks.
You could make each truck autonomous...
Or make one autonomous, and the others "followers".
This is the perfect example of a master/slave architecture.
- The leading vehicle is the master.
It is equipped with "vision" sensors and is driving like a normal self-driving truck. It's also communicating its every moves to the followers.
- The followers are slaves.
Their only task is to listen to commands the master gives, and to execute these... for example, "brake".
This type of communication is commonly called V2V (Vehicle To Vehicle), and more commonly V2X.
A lot of companies are investing in high-quality networks because they believe V2V will be the future of self-driving cars.
Vehicle Platooning is a part of the Vehicle to Vehicle family.
The family also involves multiple types of communication between vehicles that all have one purpose: reduce traffic, reduce CO2, and make the roads safer.
The 3 Platooning Affirmations
With what I just told you, we can write down 3 main affirmations.
- Platooning is mainly doable on highways, the convoy is in the right lane.
Any Platooning solution in the middle of the city will take years to be marketed.
- Vehicle Platooning can reduce fuel consumption and CO2 emission for followers, especially if the vehicles are close to each other.
Why? Because of the atmospheric drag that will be reduced.
👉 Tests estimate that at 80 km/h, we can see a 20% fuel reduction if the distance is short (a few meters).
If you read a bit about the topic, you might come across the word "HDV".
It actually means Heavy-Duty Vehicle, and it corresponds to "big trucks".
For HDV, the CO2 emission reduction (for the followers) is high, and it's where Platooning is the most interesting. To get proof of that, I encourage you to read this research paper that demonstrates this using tests.
The drag ratio is higher for buses and HDV, and lower for Light vehicles.
Finally, it also demonstrates that the closer the vehicles are to each other, the higher the reduction.
- Vehicle Platooning is mainly used for trucks; because this is where the CO2 "save" is the highest.
So here is the full picture.
Now let me tell you a story...
I once had an interview with a company that wanted me to work on their self-driving car system.
My first contact with them was a phone call.
One thing that kept boring me is that all along the interviewer kept talking about "automated cars'.
And I kept responding "It's autonomous".
Ultimately, I ended up thinking "these guys don't know what they're talking about! What the hell!"
Later in the interview, they told me about a Platooning solution they were developing.
In this Platooning solution, the leading vehicle is an autonomous car, but the followers are "automated".
I realized they were truly talking about "automated cars", and that I was the one who didn't know what he was talking about.
"Sometimes you think you know, and in fact, you just don't." (Me)
Anyway, we can already understand the fact that platoons don't need as much driverless technology as normal self-driving cars.
The main word behind their technology is Adaptive Cruise Control, which corresponds to what we already have in most modern vehicles:
- The ability to adjust your speed to the front vehicle
- The ability to brake at any time
- If desired, the ability to change lanes
The distance, and braking
One great advantage of Platooning is that the "reaction time", usually 1 second, is close to zero here.
When the leading vehicle transmits its "brake" command to the followers, the followers instantly get the message.
If something is on the road, the followers won't have to wait for an emergency moment to brake, they will be able to prepare to brake instantly.
They will be able to reduce their speed all at the same time, respecting the distance between each vehicle.
The distance between vehicles is usually set to 10 meters.
This is the distance that is both safe for the vehicles, and efficient, in terms of CO2 reduction.
The Levels of Platooning
A basic Platooning solution will make the vehicles so close that no car can join the convoy.
Sometimes, it happens, and an evolved solution should be able to manage this.
The followers must, therefore, be equipped with the level of technology to:
- Detect vehicles and anticipate their insertion in the middle of the convoy
- Let the convoy reduce its speed and intelligently integrate a new vehicle inside that may leave after.
Other V2V Solutions
If we're discussing automated cars, the master doesn't have to be a vehicle.
Imagine if a street light was equipped with a LiDAR...
It could send the 3D coordinates of every static vehicle and the exact map of the world to "automated vehicles". These vehicles would then simply be equipped with basic solutions to brake when needed. They wouldn't have to be autonomous, just automated. Why should everyone perform mapping, localization, and perception, when most results are similar?
When we're rethinking the infrastructure, anything can be possible...
👉 Vehicle Platooning will eventually be the go-to for autonomous trucking.
👉 It will skyrocket when 5G solutions become mainstream... because 4G currently has flaws, and Platooning relies entirely on vehicle communication.
👉 It's a powerful instrument against CO2 consumption.
👉 It must be designed intelligently and is potentially harder than the "self-driving car" problem.
👉 It can drastically reduce the cost in terms of fuel, and technology - we could use 1 LiDAR instead of 10... LiDAR would here be a bargain.
📩 To learn more about self-driving cars, I’m sending daily emails to aspiring Self-Driving Car engineers! Join here and learn how to break into the cutting-edge field!