Last week, Tesla announced a new hardware setup for future Model 3 and Model Y vehicles based solely around cameras instead of the previous setup that used cameras and radar sensors.

This is a step away from virtually all other autonomous vehicle producers, which are developing their hardware suites using lidar sensors – something Tesla founder and CEO Elon Musk has long shunned and called a “crutch”. So why is Tesla moving even further away from the hardware everyone else is using?

Although the new camera-only hardware will require Tesla to temporarily reduce the autonomous capabilities of their vehicles, Elon Musk has always held the belief that, since humans only use visual capabilities to drive, autonomous vehicles can do the same. He is determined to solve the more difficult software problem of processing camera data rather than relying on more hardware. There are a few reasons why he might want to go that route. For one, radar sensors are more expensive than cameras, and lidar sensors are even more so. Likewise, the data produced by radar and lidar sensors requires much more computational power. That means a greater number of – and more powerful – computer chips. All of this adds up to considerably more cost to the tune of thousands of dollars. If the median price of a new car is around $35,000, then a $3,000 hardware cost savings amounts to a nearly 10% reduction in the cost of production. That gives Tesla the wonderful dilemma of whether to pass along the savings to the customer and sell a vehicle that is priced well below the competition, or to pocket the savings and add a massive amount to their gross margins.

Of course, there is the separate possibility that the driving factor behind this decision is Musk’s own personal ambition to do something that everyone else says is impossible…

We may never know the true reason that Tesla made the decision to cut out radar, but one thing is certain: If they are able to figure out how to make it work, they have a huge advantage over their competitors.