Trade secrets, confrontation and junk science

Back in 2005, a Sarasota County (FL) judge ordered CMI, the maker of the Intoxilyzer, to produce the source code to a designated expert for examination. A judge in neighboring Manatee County did likewise. Two years ago, a Florida appeals court upheld the trial court orders. Earlier this year the Florida Supreme Court refused to hear CMI's appeal.

Today, six years later, the source code has yet to be produced.

Things reached a head last week when prosecutors made the decision to dismiss a DWI case against Janet Landrum rather than have a judge sign an order that could invalidate any prosecutions based on breath test results. Ms. Landrum was being tried after being arrested for the 11th time for driving while intoxicated.

Earl Landrum, the State's Attorney, has been CMI's water-carrier, arguing that the source code is a trade secret and that defense experts have no right to inspect it.

Florida uses the Intoxilyzer 8000 which uses dry gas calibration (rather than the alcohol jar on the Intoxilyzer 5000) and measures the absorption of light at two wavelengths (rather than five).

No one outside of CMI has seen the source code that runs the machine. Not one state's expert can testify exactly how the machine operates. Since CMI fights every attempt by anyone to see the source code, the reliability of the machine cannot be tested.

Just think about that for a second. The state will put an "expert" on the stand who will testify that the machine was working properly at the time of a given breath test. That "expert" will testify that the machine had been maintained properly. That "expert" will testify as to the alleged alcohol concentration in a motorist's blood. But that "expert" can't tell anyone exactly what goes on in the innards of that machine.

That "expert" knows how to run a diagnostic check on the machine, how to change out the dry gas canister, how to clean the machine and how to make simple repairs to the machine. But that "expert" can't tell a jury how the machine's computer program converts the absorption of light into a volumetric measurement that is supposedly analogous to the level of alcohol in a person's bloodstream.

In an era in which we have learned that the forensic "science" that was used to obtain convictions is, in many ways, junk science, we are supposed to put our faith in a grey box that no one outside CMI can explain the manner in which it works (or doesn't).

What are y'all hiding out there in Kentucky, CMI? The rights of a criminal defendant to confront and challenge the evidence arrayed against him trumps any claim that the source code is a trade secret. The mere fact that CMI is unwilling to turn over the source code to designated experts is more than enough to cast a shadow of doubt upon the accuracy of the machine.