Last May, Tesla stuck the middle finger to the entire auto industry, deeming the use of multiple sensors for autonomous driving functionalities… unnecessary. It ditched radars and initiated a camera-only approach, dubbed “Tesla Vision.”

And while we’re used to Elon’s unorthodox methods, Toyota’s decision to follow suit is definitely surprising. The car maker’s self-driving subsidiary Woven Planet will be the second company to adopt a vision-based strategy in pursuit of fully autonomous driving.

If you’re wondering WHY, I feel you. It made no sense for Tesla and nor does it for Toyota.

Cameras may offer higher resolutions, but have certain limitations. They have lower accuracy in terms of object distance and speed (especially during nighttime), they offer limited range, and they are less effective in bad weather conditions.

But… they’re cheaper, and that’s the main reason for the switch, as Michael Benisch, VP of Engineering at Woven Planet told Reuters.

Benisch highlighted the need for big amounts of data, which the company can’t collect from its currently small fleet of “very expensive” autonomous vehicles.

To solve this, Woven Planet is opting to use a larger fleet of vehicles, equipped with cheap cameras, to collect data and train its neural network system, hoping to drive down costs and accelerate its self-driving efforts.

Still, cameras are less reliable than radar and LiDAR technology, which questions if the data provided by them is good enough to make future self-driving cars safe.

Plus, the neural networks that determine what’s being seen by the cameras call for large amounts of training and processing power — and both of these factors are intrinsically limited within a car’s computer system, for now at least.

In comparison, Tesla’s camera-only approach hasn’t been without issues.

First up, it cost the company the safety recognition from the National Highway Traffic Safety Administration (NHTSA).

It also led to multiple owner complaints about “phantom braking” (unexpected activation of the brakes) issues, possibly related to camera vision inaccuracies. This resulted in investigations and recalls by the NHTSA.

At least, Toyota isn’t planning to apply this strategy to its commercial vehicles. It’ll still use multiple sensors including radars and LiDAR for its robotaxis and other autonomous vehicles on the road, as this is currently deemed the safest approach.

“But in many, many years, it’s entirely possible that camera type technology can catch up and overtake some of the more advanced sensors,” Benisch said.

Let’s hope that until then Toyota won’t act like Tesla.