Tesla CEO Elon Musk is not a fan of LIDAR - the laser sensor used by many autonomous

car companies to navigate the world.

LIDAR is a fool's errand and anyone relying on LIDAR is doomed.

LIDAR stands for Light Detection and Ranging.

The sensor emits pulsed light waves into the environment.

They bounce off surrounding objects and return back to the sensor.

The sensor uses the time it took for each pulse to return to the sensor to calculate

the distance it traveled.

Repeating this a million times a second helps create a 3D map of the environment.

If the LIDAR detects it's getting close to an object, it'll apply the brakes to

slow down the vehicle.

Practically everyone other than Tesla views LIDAR as essential.

Alphabet's Waymo – seen as one of the frontrunners of self-driving - uses LIDAR.

GM's self-driving unit Cruise acquired its own LIDAR manufacturer a few years ago.

Apple's rumored smart car is said to be looking for the sensor.

One reason Musk doesn't like LIDAR is that it's expensive.

It's like having a whole bunch of expensive appendices.

One appendix is bad. Well, now we're going to put a whole bunch of them.

That's ridiculous.

You'll see.

LIDAR used to be very expensive, costing tens of thousands of dollars.

But now, LIDAR units can be thousands of dollars or under $1,000.

And manufacturers aspire to reduce costs to the low hundreds in the future.

Supporters of LIDAR argue it can do things cameras cannot.

While cameras interpret depth, LIDAR precisely measures distances.

It's trusted by NASA to do everything from studying changes to glaciers to collecting

data over the Amazon rainforest.

However, Musk argues that LIDAR cannot fully capture the quirks of the real world.

He's called it a crutch.

Whereas Tesla's approach to self-driving is to rely on cameras in the way humans depend

on vision to drive.

To use just computer vision, it would have to understand an image in the way we understand

an image with our eyes and brain.

Musk thinks it can be done.

Which begs the question: Why was Tesla spotted with LIDAR?

An autonomous vehicle consultant recently posted these photos on Twitter of a Model

Y that was sporting LIDAR in Florida.

According to a report by Automotive News, Tesla has a contract with LIDAR manufacturer

Luminar Technologies though neither Luminar nor Tesla has confirmed it.

I don't think that they'll ever actually use LIDAR on their production vehicles.

I spoke with Sam Abuelsamid, an analyst with Guidehouse Insights, who believes Tesla is

simply using LIDAR for training purposes.

If your machine vision system is saying there's a vehicle 50 m away from you in the lane to

your right, you want to have something to use to verify that the software is calculating

that correctly, and that's what LIDAR does very well.

He says if Tesla were to change its mind and incorporate LIDAR, that would be problematic

for its fleet of a million existing vehicles.

Musk assured customers that their Teslas already had all the hardware necessary for full self-driving.

In the same way Tesla feels it doesn't need LIDAR, it's also doing away with radar.

Tesla's website used to refer to radar for its driver assistance system, Autopilot.

However, the references to radar have recently been dropped.

The company confirmed Model 3s and Model Ys delivered to the North American market are

no longer equipped with radar.

Model S and Model Xs will eventually make the transition too as will cars destined for

other markets.

Tesla says new vehicles without radar will have Autopilot temporarily limited to lower

speeds and a longer minimum following distance and will lose a few other functions.

Removing radar isn't surprising.

Musk tweeted in 2018 that “Long-term, the car will work purely on vision”.

The company is going all in on what it calls “Tesla Vision” where Autopilot and full

self-driving - its most advanced driver assistance system - will rely purely on camera vision

and neural net processing.

This is what Tesla Autopilot sees using neural networks that analyze the vehicle's surroundings.

The aim is for Tesla's artificial intelligence technologies to propel its system from Level

2 autonomy all the way to Level 5 in which a driver would become a passenger.

There would be no steering wheel or brakes.

Tesla believes that is possible without radar or LIDAR though some analysts question whether

that's the best move.

Although Algolux is a company that focuses on using cameras for autonomous driving, an

executive there tells me relying solely on cameras may not be the best option:

Dave Tokic says “…it may be short-sighted to not leverage additional sensor modalities,

especially new breakthroughs in radar…as well as Lidar if the market can absorb it.”

Despite his adamant belief that LIDAR will be unnecessary, Musk once admitted: Perhaps

I am wrong, and I will look like a fool.

But I am quite certain that I am not.

Tesla is relying heavily on artificial intelligence as it prepares for a driverless future.

There's a fascinating class on Skillshare about how machine learning and AI are shaping

our world.

My sponsor Skillshare offers thousands of inspiring classes on a whole host of topics

- which you can try out for FREE.

MKBHD has great tips on how to grow a YouTube channel.

Polymatter has a class on how to make animated videos.

Each class is broken up into short lessons, there are no ads, and it's for all skill

levels.

The first 1,000 of my subscribers to click the link in my description will get a complimentary

trial of Premium Membership so you can explore your creativity.

For Newsthink, I'm Cindy Pom.