The problems surrounding law enforcement’s ability to detect impaired driving resulting from cannabis use are becoming more apparent as more states legalize the drug. Michael Milburn, chief science officer and founder of Impairment Science Inc says his invention, the Druid app, could solve the problem. We asked him for details on this cutting edge technology.

The Paper.: Can you explain what Druid is and what it does?

Milburn: Druid is an app that measures cognitive motor performance and impairment. It’s grounded in cognitive neuroscience. We measure reaction time, hand-eye coordination, balance and ability to perform divided-attention tasks—which are, of course, critical for driving or operating heavy machinery, or anything like that. It takes just about three minutes to perform the test. Shapes—circles or squares—flash on the screen, and you have to touch the screen in different areas.

We were awarded a Small Business Innovation Research Grant from the National Institutes of Health to do research. We funded research at Johns Hopkins that demonstrated that the app can distinguish different levels of impairment—from different amounts of THC that a person consumes. No other tool in the world can do that that I know of.

But it’s not just a cannabis app; it is a cognitive motor performance app. It can measure if a person becomes impaired when they consume alcohol or drugs or cannabis, or if they’re highly fatigued. Alcohol and drugs aren’t the only source of impairment. There’s illness, injury, chronic medical conditions, fatigue. Druid can pick up increases in impairment from any of those sources.

How does it work?

There’s four different tasks. For the first task, shapes appear on the screen—a circle or a square. You have to touch the screen in a different place depending on where the shape lands. So your reaction time is measured. In the second task, you’re estimating the passage of 30 seconds while you’re also hitting the shapes that flash on the screen. For the third task, a circle on the screen starts moving around and you’re tracking it with your finger while squares are flashing on the screen that you have to count. And then the fourth task is the single leg stand from the standard field sobriety test. That’s an important measure because balance is controlled by the cerebellum. And a half to a third of the neurons in the brain are in the cerebellum, so when you have disrupted balance you have system-wide impairment.

The app generates a few hundred data points and we statistically integrate them into a single score that is reliable and validated. We’ve actually been able to calibrate our scores. They go from zero to 100. The higher the impairment score, the more impaired the person is. We’ve been able to calibrate it to the impairment levels for blood alcohol concentration.

How does it compare to the more traditional field sobriety test?

Well the traditional field sobriety test is completely subjective. Officers are trained to spot different cues, and I like that, but it’s totally up to the officers judgment. The druid app, in contrast, is an objective measure. The accelerometer in the device measures the amount of movement and instability when a person is doing a single leg stand. The reaction time is measured in milliseconds. Research has shown that the standard field sobriety test is not effective in assessing impairment for marijuana. For the Druid app—we have multiple peer-reviewed validation studies published on its effectiveness at detecting alcohol and cannabis impairment.

We just completed a big study on fatigue in Texas for a police department. They were looking at potentially scheduling different shifts—scheduled 24 on and 48 off, or 48 on 94 off—and we were able to forecast the fatigue officers would experience during a shift.

Why is it more important for officers and employers to test for impairment levels rather than the presence of drugs or alcohol?

Drug testing is a really flawed approach. Drug testing does not measure impairment. People develop tolerance to drugs, and simply measuring how much cannabis a person consumes is not necessarily going to tell you what their level of impairment is. It’s important to measure actual behavior and actual impairment, not just the amount of drugs in the system.

Have any law enforcement agencies shown an interest in using the app to detect impairment?

Yes, actually I’m going to be speaking to the Colorado Impaired Driving Taskforce in a few weeks about Druid. And I’ve actually worked with some cadets at police academies in Massachusetts, where they run what’s called “wet labs”—where they bring in volunteers who drink beer to get their blood alcohol content up to over 0.8 percent, and then the cadets can practice the standard field sobriety test on actual drunk people. We gave people the test when they were sober, and then they had to start drinking. Druid was able to show a massive increase in impairment.

We’re certainly getting some attention from law enforcement, but the roadside testing is realistically a few years down the road. Because the politicians, the courts, scientists, law enforcement—everyone wants their say in this particular arena.

The app can also be used by drivers who want to check their levels before getting behind a wheel?

As I’ve said for years, I didn’t create druid to bust stoned drivers. I want to stop them from getting in the car in first place. In the pre-Druid era, there was no way for someone to actually tell how impaired they were. If someone’s been smoking joints for a few hours, and it’s time to go home, they might say, “I think I’m okay to drive.” In contrast, there’s now an objective measure of whether they’re okay or not.

We’ve actually signed a contract with Grupo Modelo, which runs the largest brewery in the world in Zacatecas, Mexico. They have purchased 20,000 licenses for Druid downloads, and they’re going to distribute them with QR codes at roadside stops in the town of Zacatecas. They’re trying to reduce harm from impaired driving, so we’re really hopeful that we can demonstrate a real societal impact.