What was only a science fiction story a few years ago is rapidly approaching reality. Autonomous driving -- that old concept of self-driving cars or SDCs -- has grown into an utterly serious field of research for multiple companies, including Google.

Taking a long-term perspective, car manufacturers are already locked in competition over who can adapt the best autonomous technology.

The timetables vary, of course. In 2013 Nissan declared that the first self-driving cars would be hitting the road around 2020 (although autonomous cars may be end up being much easier to release in Nissan's home country of Japan compared to the vast roadways of the U.S.).

An early 2014 study predicted that by 2035 self-driving cars would be common, at least when it came to early models that require a co-driver who had to manage some minor tasks.

Meanwhile, Ford is at work with institutions like MIT and Stanford to create the right network of sensors to make a real-time map that Ford cars could use to navigate tricky routes.

A quick look at CES 2014 cements the notion that SDCs are only a few decades away, at most, thanks to the combined efforts of the tech giants, excited universities, and investment-happy manufacturers. The question is, who will actually release the first mainstream models, and where they will be?

For now, it's more likely that manufacturers will release models outside of America -- in Japan or Europe first -- and stay away from the complex and lawsuit-happy arena of U.S. highways for as long as possible. However, some states such as Nevada are already working on the first laws and testing requirements for prototype self-driving cars, in preparation for the opening of the market.

Drive me!: What self-driving cars could mean on the road

When autonomous vehicles do hit the streets, they could change everything about how we drive. The necessary skills and licensing for SDCs would naturally look a lot different from today's "manual" cars. Tech and car companies are swift to promise safer roads with minimal collisions and reduced danger, but the truth is that a lot of questions still have to be answered:

1. Shifting laws

How will SDCs meet the necessary rules of the road? This goes far beyond a bit of basic programming, because driving rules are ever-shifting, complex matters. Different countries drive on different sides of the road; that alone requires an entirely different software and sensor set-up.

image via shutterstock

Even within a single country like the United States, driving laws vary considerably from state to state. Will SDCs change behavior based on what state they are sold in? Will they have to stop and enter (wireless?) service whenever they cross into a new state? What does this mean for people who commute between state lines every day?

2. Impatience

We are used to driving ourselves -- at the exact speeds that we want, when and where want. Autonomous driving takes some of that control away, and it's not yet clear that we are ready for that.

Yes, laptop and smartphone work in the car sounds great if you're stuck in a freeway line. But what if you just want to make a quick run down to the grocery store before dinner at speeds a bit above the posted limit? A fully automated car would not allow it, and that could lead to a lot of frustration.

3. Personal liability

How will state laws change regarding personal liability in accidents or drunk driving? At the moment, traffic tickets and DUI records add up to an important profile of driving behavior that courts depend on. SDCs could destroy or revolutionize this precedent-based system.

4. Inaccessible regions

No matter how good SDCs get, there will be some places they can't go. Bad winter storms, lots of snow, monsoons, unmapped dirt roads ... all require manual driving. The result will be a massive market split between the safe urban areas that "work" for autonomous cars and the other areas where manual cars are still marketed and driven.