Ken Fichtler is the founder and CEO of Gaize, a first of its kind real-time impairment detection platform. Gaize uses analyzes eye movement changes using artificial intelligence to discover impairment from cannabis and other drugs. Here is Fichtler’s take on the current state — and future — of the industry.
Can you give us some background on the origin of Gaize and its specialization?
While I was serving as the Director of Economic Development for Montana, I spent some time working to understand the impacts of cannabis legalization. Montana was a medical cannabis state at the time, so I wanted to know what recreational cannabis would change. As I talked with people in states that had recreational cannabis, the one problem that kept coming up was the lack of a device that could measure cannabis impairment. This fascinated me and struck me as both a huge safety problem and an attractive market opportunity. I studied the problem for the next few years and eventually landed on the solution we’ve now built.
Basically, since THC lingers in the body for long after it was last used, you cannot simply measure THC content in the body to determine if someone is currently impaired. THC therefore behaves much differently in the body than does alcohol. While learning about how impairment is measured today, I found that the police use specially trained officers called Drug Recognition Experts. They conduct a series of tests to determine if someone is experiencing impairment for the purposes discovering and prosecuting of impaired driving cases. In this examination, officers try to elicit certain signs of impairment that manifest in the eyes. These eye tests in particular seemed to be both accurate and thoroughly studied. The problem with these tests is that they are conducted by highly trained humans, so that won’t scale enough to solve the challenge of detecting impaired drivers and workers.
What struck me was that if we could automate these eye tests using a VR headset and use eye tracking sensors to capture data, we should be able to detect the signs of impairment even better than a human.
Describe the device and how it works. Is it uncomfortable for the person being tested?
Gaize uses a special VR headset with embedded eye tracking sensors to conduct a fully automated version of the Drug Recognition Expert eye tests. The product captures high-precision eye movement data and video. The data is analyzed using machine learning models that detect the same signs of impairment that Drug Recognition Experts currently look for. We provide the video as a piece of human-interpretable evidence.
The test is pretty boring overall. It’s definitely not a fun VR experience and is simply a ball that moves on the screen in the patterns specified by the law enforcement training manual. We also use varying levels of light and measure how the pupils respond. There is no VR sickness when using Gaize because there is no dynamically moving environment.
Living in California, where cannabis is legal, there have been a lot of conflicting information when you try to determine if cannabis will prevent you from passing an employer-mandated drug test. Tell us how CA AB2188 will address that.
When California passed AB2188, it was the nation’s strongest law that protects the rights of cannabis users, while also providing a pathway for employers to maintain a safe workplace. THC and its metabolites often linger in the body for over a month since cannabis was last used. Many scientific studies have tried to establish a correlation between how much THC is in the body and the level of impairment being experienced. Unfortunately, those studies have universally shown that there is no such correlation, so measuring THC is only useful in showing that someone has used cannabis previously. Because of that, employers taking adverse action against employees for having THC in their body is like telling an employee that they can’t have a beer after work.
CA AB2188 rectifies that by specifying that employers cannot use THC metabolite tests to take adverse action against an employee. Instead, employers need to prove that an employee was impaired on the job.
How will Gaize help employers maintain compliance with CA AB2188?
On January 1, 2024, employers must be in compliance with AB2188. That’s just a few months for businesses to make often substantial changes to their drug testing policies. Gaize is an AB2188 compliant test that shows only active impairment from cannabis and other drugs. It’s an easy-to-use and non-invasive solution that allows employers to maintain a safe workforce and respects employee rights.
What would the distribution look like? Would you sell to labs that currently do testing or to the employers themselves? Or both?
Both. We currently have customers in almost every industry with safety sensitive employees and we also sell to drug testing companies. These companies do Gaize tests as a service for their customers, and can conduct any follow-up testing as well.
You did the world’s largest cannabis impairment trial. Can you frame what you were wanting to prove and what you found?
Yes, the clinical trial was intended to provide training and validation data for the Gaize product. We brought 350 people into a facility in a sober state, measured their eye movement, and then got them high using recreationally available cannabis and measured their eye movement at four additional times. This ended up being the largest clinical trial investigating cannabis impairment, and the resulting dataset is the largest impaired eye movement dataset in the world. The objective was to find signs of impairment in eye movement, and build machine learning models that can detect those signs. We were able to so, and validate the effectiveness of our algorithms in detecting the signs of impairment with greater than 95% accuracy.
How was this device developed?
Gaize is a complex product with a hardware component and several software applications that all work in concert. We use an off-the-shelf VR headset for Gaize, which radically simplified our path to market. The same headset has been used in several medical devices that use eye movement to detect medical issues, so we were confident in its capabilities from the outset and it has proven to be an outstanding product. To develop the software applications, we used both in-house and contracted developers.
You also sell the device to law enforcement. Do anticipate wide adoption?
Yes, the reception from law enforcement has been good so far with several pilot deployments in key cities happening now. We’re doing something that requires law enforcement officials to rethink processes and challenges the status quo. So, while we’re very confident that the right people are looking at the product and it’ll be deployed broadly soon, we don’t have an exact date at this time.
Can you scale quickly if adoption happens very fast?
Yes, within reason. We have the ability to supply several thousand units per month currently, and can scale beyond that within a few months if required.
Can this test for other impairment?
Yes, Gaize currently detects the same signs of impairment that law enforcement look for each class of drug. We’re working toward being able to classify the impairment we see though. The vision for Gaize is to create a singular device that can detect and classify impairment from any drug. We wanted to solve the cannabis impairment issue first and foremost, and have built algorithms that do that with excellent accuracy. Today, we’re moving on to other drugs. Our team has already captured data on alcohol, and we’re now working on both opiates and stimulants.
Copyright © 2023 California Business Journal. All Rights Reserved.