I'll speak to both the false positives and the malfunctions. The malfunctions were articulated in the report that we wrote. However, a lot of them dealt with things like the printer not connecting to the Bluetooth, so they weren't really malfunctions with the devices themselves. Overall, they worked very well for the officers.
With regard to the false positive rate, because this was a volunteer pilot project, we assumed—because the officers screened the individuals before they actually volunteered for the pilot—that there were a few false positives. The device recorded the same drug detection in more than one instance in a row. Overall, the individuals who were part of the pilot were screened by the officers and didn't stay after the fact for us to do any further investigation about their consumption of a drug.