Should self-driving cars use radar or laser as their eyes?

You’ve likely heard about it. Most auto manufacturers are working on some sort of solution for autonomous driving. As with many new technologies, there are inevitably different approaches to solving the same puzzle. Those of us who grew up in the ‘80s will remember the fight between Beta and VHS for home video dominance.

If that audio-visual battle royale taught us anything, it isn’t always the best idea that wins.

The same can be said for the current efforts being expended to develop cars that can drive themselves. We’re not to full autonomy yet, but advances are being made. Tesla, the company that lives on the cutting edge of technology, thinks it can crack the code using radar. In the opposite corner, most other car makers believe LIDAR is the answer. Who’s on the right track?

A quick Radar/LIDAR tutorial.

Before diving in, let’s convene a quick science lesson. Radar is actually an acronym for radio detection and imaging. Radio waves (pulsed or continuous) emitted from a transmitter reflect off objects and return to the receiver, giving information about the object’s location and speed. This is traditionally how roadside cops figure out you’re exceeding the speed limit, for example.

LIDAR is also a surveying method. It, however, calculates the distance to an object by illuminating the target with pulsed laser light and measures the reflected pulses with a sensor. Differences in laser return times and wavelengths can then be used to make digital 3-D representations of the target. A modern machining tool may use lasers to precisely measure an object in order to duplicate it.

Okay, science class is over. This still leaves us with the question: which is better?

Which one wins?

Radar knows the speed of objects it ‘sees’. It is also cheap, robust, and easy to build into the front of a car. The thing is, it also detects many items a car needn’t immediately worry about, such as directional signage or stray hubcaps. Developers of radar for autonomous cars therefore need to make a difficult choice: program the system to ignore these items and focus on the stuff that’s moving – namely, other cars on the road.

To be clear, Tesla combines the eyes and ears of two other systems – distance measuring Ultrasonic and Passive Video – to facilitate its Autopilot.

LIDAR, which stands for light detection and ranging, has become a common feature of autonomous rides operated by companies like General Motors and Waymo. These sensors use lasers to build a very precise and detailed map of the world around a car. Because it does so, it can easily distinguish between a stray hubcap and another car. This is beneficial to the car’s occupants since there are many stationary objects that can appear in a roadway and hinder forward motion.

Here’s the problem: compared to radar, LIDAR is very new technology. It is eye-wateringly expensive and, in most cases, is not robust enough to survive a life of road hazards and bad weather. It also takes several cubic yards of cash to develop for an individual model; given that most manufacturers have dozens of models. Well, you see the issue.

Most car companies that are choosing to invest in the development of a self-driving technology are including LIDAR as a key component of the systems they are creating. The motto espoused by Tesla CEO Elon Musk, by contrast, is “not to let the perfect be the enemy of the better,” meaning they do not wish to withhold imperfect driver assistance features that nonetheless improve safety.

So, which is right?

Radar is robust and ready now, while LIDAR is more precise but not yet fully baked. Similar to the battle between Beta and VHS, there will likely only be one winner.

Got a car-related question?

Our Consumer and Technical Services (CATS) team receives questions about car care, buying a new or used vehicle, auto repairs, vehicle inspection, driving costs and more. If you have a question for CATS, call 1-866-464-6448 or email: cats@caasco.ca.

Written by: Mathew Guy