2. The machine records all observations and outcomes over time. The machine builds a database of information using every observation, every reading of the vehicle’s sensors and every decision by the driver.

3. The machine generalizes the data to make intelligent decisions in new situations. Based on all past situations, the machine can now make highly intelligent decisions — even when it encounters new situations on the road that haven’t been previously observed.

Related: Autonomous Cars Quickly Come to Life

4. Over time the machine gets smarter. The more data that are available, the more inclusive the machine can be — essentially learning from itself.

Yet, as remarkable as autonomous vehicles are, the technology is actually quite simple. Why? Because there is a finite number of decisions a driver can make in any given situation. As a result, the car uses basic data-driven analysis to detect good or bad behavior and adapts accordingly, becoming “smart” in the process. By combining that simple process with big data, autonomous cars may be on the cusp of turning a century-old dream into reality.

This article has been republished with permission from Invesco Powershares.