"We used machine learning to 'teach' the breathalyser to recognise the presence of THB based on the electrical currents' recovery time, even when there are other substances, like alcohol, present in the breath.
"Creating a prototype that would work in the field was a crucial step in making this technology applicable," said Ervin Sejdic, PhD.
"It took a cross-disciplinary team to turn this idea into a usable device that's vital for keeping the roads safe."
As weed becomes legalised or decriminalised in more and more places across the globe, law enforcement will need a device like this to test if a driver is under the influence, much like they test for alcohol.
The device is said to be a promising first step in this direction.