???? The Intoxilzyer 5000 EN, which is the only post-arrest breath testing device approved for use in New Hampshire, has many parts which have been studied by engineers, scientists, forensic breath testing experts and courts in many states.? The result of all this study has been a rich literature detailing the situations in which the results spit out by this machine may not accurately reflect deep lung breach alcohol concentration (BrAC).
?? But there is one essential part of the Intoxilyzer 5000 which has been off limits to study–the computer source code which runs the whole machine.? Although the machine uses infrared spectrometry to analyze a fixed sample of breath, it is the computer software that is responsible for:
- Telling the machine what numbers to report;
- Telling the machine whether it can report a result at all;
- Correcting for possible interferents (e.g. chemicals other than ethyl alcohol);
- Correctly reading the machine’s internal and external standards.
??? The machine’s so-called “slope detector”–which is supposed to tell the machine to abort when BrAC continues to rise steeply when the subject is blowing, due to the likely presence of mouth or stomach alcohol–depends entirely on the source code.? I’ve heard a number of highly qualified experts say that the slope detector does not always work correctly.? I’ve also heard a world class expert say that the?machine?often reports inaccurately high BrAC results when?the subject has a small amount of alcohol (the proverbial two beers) but has been exposed to potential interferents (certain volatile chemicals that mimic alcohol’s infrared spectrograph).? Although the software is supposed to correct for?such interferents, it may not always work correctly when there is?some alcohol in the mix.?
???? So how do we know that?the Intoxilyzer 5000’s source code is doing its job correctly?? As a result of some dogged litigation by DUI defense lawyers in Florida, Kentucky, Minnesota and New Jersey, the Intoxilyzer’s source code will finally be analyzed by qualified computer scientists.??
???? More after the break.?
???? In recent years, DUI defense attorneys began to request copies of the source code for the Intoxilyzer 5000 and other machines.? Eventually a number of?trial and appeals courts?ordered the production of the source code.?
?? In New Jersey–where a team of attorneys and experts doggedly pursued the issue to the state supreme court–the source code for that state’s machine (the Alcotest 7110 manufactured by Draeger Safety Diagnostics, Inc.) was produced and analyzed.? The forensic source code experts determined that entire machine was unreliable because the source code was bug ridden and could not pass industry standards for software development and testing.? They also determined that the Alcotest’s source code actually disabled the machine’s catastrophic error detection software–so that the machine will spit out BrAC results when it should instead abort!? As the expert report states, “turning off these safeguards means as these conditions are encountered the machine produces unpredictable results.”? The experts hired by the?the Alcotest’s manufacturer agreed that the source code was?”not written in a manner consistent with usual design?best practices” but concluded that “there are no obvious defects intentionally written to produce anything other than consistent results.”?The New Jersey Supreme Court will soon determine whether (a) the Alcotest’s source code is sufficiently reliable to allow any continued use of that machine in court and administrative license suspension proceedings and (b) if so, under what conditions the machine can be used.
?? The New Jersey experts’ findings have no direct applicability to New Hampshire because we use?the Intoxilyzer 5000 EN, not the Alcotest 7110.? To date, no New Hampshire court has ordered the production of the Intoxilyzer’s source code and, to my knowledge, no defense experts have ever examined the source code in any jurisdiction.? However, the Intoxilyzer’s source code has been the subject of highly contested litigation in at least three states.
??? In Minnesota the courts have ordered the state to produce the source code in hundreds of DWI and ALS cases.? The Intoxilyzer’s manufacturer, CMI of Kentucky, Inc. refused to produce the source code, claiming that it was a trade secret and that it was not in the state’s control.? The DUI defense attorneys convinced the Minnesota courts that, pursuant to its contract with CMI, the State of Minnesota did in fact have the right to the source code and, therefore, the state could be ordered to produce it.? The courts also held that, while CMI would be entitled to a reasonable protective order to safeguard its legitimate interest, it could not require either the state or the DUI defense bar to sign the ten page confidentiality agreement that the company proposed.? Last week–with hundreds of DWI and ALS cases facing dismissal, the State of Minnesota sued CMI in federal court seeking to recover the purchase price it paid for?Intoxilyzers?it can no longer use until the source code is produced.? If you’re interested, the Complaint in the federal lawsuit in Minnesota can be found here.
???? In Florida, there have also been dozens, if not hundreds of cases in which accused drivers have sought access to CMI’s source code for the Intoxilyzer 5000.? While the Florida courts never ordered the state to produce the source code, a number of trial and intermediate appeals courts?ordered CMI to produce the source code pursuant to interstate subpoenas.? At first, CMI refused to honor these subpoenas claiming that?the Florida courts lacked jurisdiction.? Then CMI insisted on onerous confidentiality agreements that would have stymied rather than assisted the Florida courts.? Finally, however, CMI appeared through counsel in a Florida court and agreed to produce the source code if a protective order was in place.? Two weeks ago the Florida court ordered CMI to produce the source code and issued its own reasonable protective order.? The Florida decision is here.
?? Finally, in late 2007, a Kentucky court of appeals ordered CMI to produce its source code to DUI defense experts.? The court’s opinion is here.? Since CMI is a Kentucky corporation with substantial assets in the state, it will have no jurisdictional defenses and no ability to avoid compliance with the Kentucky order (unless, of course, it is reversed or overruled by statute).
?? So, why has there been no source code litigation in New Hampshire?? Well, any DUI defendant who is fortunate enough to get the source code would have to spend tens of thousands of dollars for an expert to analyze it.? Chances are that, despite the increasing number of appellate court decisions rejecting CMI’s trade secrets claim, a New Hampshire DUI defendant would have to litigate the issue at both the trial and appellate levels.? That would be very costly as well.? Finally, the holy grail of source code litigation–a hearing to determine whether the software is reliable–would entail many days of testimony by a group of high priced experts.? Whatever the result of that hearing, it would be appealed by one party or the other the New Hampshire Supreme Court.? The?whole endeavor would take several years?(as was the case in New Jersey, Florida, Minnesota and Kentucky) and cost somewhere in the six figures.?
??? Of course, there is always the chance that a DUI defendant could?get an order?requiring CMI to produce the source code and then an order of dismissal when the company refuses to comply.? The State might also choose to plea bargain some source code cases on very advantageous terms.? Nonetheless, if you ask for the source code you have to be prepared for the possibility that you might just get it.
?? Chances are that the issue will either be raised by a group of attorneys pooling their efforts in a number of cases with?some pro bono assistance from both attorneys and experts.? Failing that, however, New Hampshire may just have to wait to see how things shake out elsewhere once the source code is analyzed.
?? As for me, I’d be happy to demand the source code and seek a hearing on its reliability if I had a client who wanted to start down that track.
