Let’s take a moment away from all the election drama to look at what’s quickly becoming a sticky, complicated law enforcement challenge in the United States. With the number of states allowing either medical or recreational marijuana use, we’re realizing that determining when someone is too impaired to drive is a bit more complicated than in the simpler times of just checking someone’s blood alcohol levels. This poses a technological challenge where engineering hasn’t caught up to the new laws on the books in many states. This article at Route Fifty examines some of the details.

As legalized marijuana becomes more common in the U.S., the number of drivers testing positive for it has been on the rise in recent years, with pot even eclipsing alcohol as the most commonly detected intoxicant in people behind the wheel, according to some research.

But this in itself does not mean there is a wave of stoned driving hitting the nation. That’s because people process marijuana differently than beer, wine and liquor, and its presence in a driver’s blood or saliva doesn’t automatically indicate that the person is impaired.

And while there are decades of research showing how increases in blood alcohol levels correspond to a person’s ability to drive safely, the same is not true for the main component of marijuana that causes a person to feel high—delta-9-tetrahydrocannabinol, or THC.

As the author notes, checking drivers to see how much they’ve had to drink and whether or not they are impaired is a science we’ve developed to a workable level for some time now. Of course, to be honest, we’re still not all that accurate in that regard either. Different people process alcohol at varying rates depending on body mass and other factors, and not everyone is as “impaired” as somebody else based solely on their BAC levels.

But it’s at least a relatively linear scale. The more you drink, the more impaired you become. As your body flushes out the alcohol from your bloodstream, the levels drop and so too does your level of impairment. That’s not really true for marijuana. People who smoke pot have THC (tetrahydrocannabinol) in their system for a long time. Just ask anyone who has failed a drug screen test. And those levels last far longer than the amount of time they are high. So how do we know if a smoker is too high to drive?

In Washington state, they are training cops to draw blood samples from drivers at the side of the road rather than taking them to a medical facility. Pardon my saying, but that sounds like a terrible idea. The training to become a phlebotomist is rather extensive and not part of normal police training. A few seminars for the cop doesn’t make me all that comfortable with them finding a vein using a flashlight on the shoulder of the road at two in the morning.

We’re going to need a lot more data before we come up with any hard and fast rules. Simply relying on the honor system isn’t going to work because people lie to the cops about how much they’ve had to drink all the time. It’s safe to assume that pot users will do the same. But impairment from using marijuana is real, so you can’t just ignore stoned smokers. Taking away people’s licenses for smoking if they’re not impaired, however, isn’t a viable option either. If we’re going to have legal, recreational pot use around the country, a new approach is needed.