Device May Aid Drunk-Driving Enforcement : $600 High-Tech ‘Flashlight’ Has an Odor Sensor

The motorist had consumed five or six beers in the two hours or so before he got into the car to drive home, but, he figured, no problem , he could maintain a steady speed and find the center of the lane well enough to avoid detection by police.

Even if he was stopped, he knew he was speaking coherently, thinking halfway straight and could probably bluff his way through the questions of a police officer. Remember, friends had told him, always say you just had two beers. That way, you are conceding the obvious--that you’ve been drinking--but not admitting you’re drunk.

And he knew the odds were--and historically have been--on his side. By many police and highway safety estimates, only about one of every 2,000 drunk drivers on the road at any one time will be arrested. One thing the police have long needed, many officers say, is a gadget they can stick inside a car to quickly detect the odor of alcohol at the roadside.

The motorist spotted red lights behind him and pulled over. The police officer poked his flashlight inside the window to scan the license. But the driver noticed a bulge on the end of the flashlight and heard a faint sort of pumping noise as the cop manipulated some small switches that, obviously, had nothing to do with turning on the light.

The cop glanced at a little three-digit readout on the top of the flashlight and then said the words the driver had always been afraid he might someday hear: “Sir, could you step out of the car?”

Possibly within a year, drunk-driving enforcement in many parts of the country--including, perhaps, some or all of California--may undergo just such a high-tech revolution with introduction of the fancy $600 flashlight. The fat end of the unit conceals a small but sophisticated sensor that detects the odor of alcoholic beverages inside a car or on the breath of the driver or other occupants.

This is no ordinary new commercial product. Instead, it has been developed under a research program set up and paid for by a safe-driving think tank supported by the nation’s automobile insurance companies. The sensor’s prospective manufacturer is not directly involved in promoting it.

The device has been tested by at least two police departments and the one that used it most extensively says it is ready to buy it now that the sensor is about to go into full-scale production and be offered for sale to departments across the country.

The Charlottesville, Va., police department, which has the most experience with the prototype sensor, found it sniffed out 68% of drivers--at a series of drunk-driving roadblocks last year--whose blood alcohol levels were .10% (California’s official minimum to establish drunkenness) or higher while officers working without the sensor could detect only 45% of the drunks. For people with just under the legal minimum--drivers with blood alcohols of .05 or more but who probably still were not sober enough to react quickly in a high-speed emergency--the sensor fingered 45% while the officers detected just 24%.

Moreover, anticipating a possible legal tangle over whether using the sensor violates civil rights guarantees against unreasonable search and seizure, the federal government’s National Highway Traffic Safety Administration has conducted an analysis that concludes the gadget passes constitutional muster.

And the head of the criminal division of the California Attorney General’s office agrees, noting--though, he emphasized, only on the basis of a description of the device by a reporter--that the sensor would apparently be legal in this state. Some police officials, however, worry that it may run afoul of unusual wording in the state’s drunk-driving laws that require a suspect to provide only one official sample of his breath, blood or urine.

While the legal questions remain unresolved, street cops seem generally excited about the new device. California Highway Patrol Officer Rick Stevens, for instance, took a look at a prototype of the device recently as it was being demonstrated by Brian O’Neill of the Washington-based Insurance Institute for Highway Safety, which paid for the sensor’s development. O’Neill was in the state to attend a traffic safety conference at the Anaheim Convention Center.

Stevens watched with interest as O’Neill manipulated the sensor, then said, “I’ve been waiting for something like this for 10 years.”

The bulging flashlight houses what is called a passive alcohol sensor. In many ways it is a sort of super-miniaturized version of the familiar breath analyzer. Since it doesn’t focus on a breath exhaled from deep inside the lungs by a suspect, however, the hand-held sensor can’t make accurate readings of the precise alcohol levels in the bloodstream. Its reading, alone, would not be enough to convict someone arrested for drunk driving.

But, O’Neill said, within 10 to 15 seconds the sensor can tell a police officer whether there is enough alcohol on the breath of a suspect to warrant closer attention and a full-dress breath test or a check of blood or urine. In a demonstration last week, the sensor had no difficulty focusing on the breath of a reporter whose total intake of alcoholic beverages in the 24 hours before the test amounted to half a mouthful of beer, consumed about five minutes before the sensor was used.

Hole in the Casing

When the sensor is turned on, it pumps a small quantity of the air around it through a small hole in the flashlight casing. The sample passes across an electrochemical fuel cell that measures its alcohol content. The measurement appears as a digital display, expressed in three decimal places--or within one-thousandth of a percent.

The prototype being circulated in the United States now was developed for the Insurance Institute by Lion Laboratories Ltd., a Welsh firm, and Prototypes Inc., headquartered in Maryland. O’Neill declined to say how much money the institute had invested in the program.

Still other types of sensors--utilizing the same technology--are either in use or under development in other parts of the world and a Japanese version, developed by the Honda automobile and motorcycle manufacturing firm, has been tested by the Washington Metropolitan Police with only limited success.

Lt. A. E. Rodenizer, who headed testing of the sensor conducted by the Charlottesville police, said the sensor, overall, seemed to have doubled the effectiveness of officers working on drunk-driving roadblocks. In a telephone interview, Rodenizer predicted that, unless use of the sensor is successfully challenged in court, the device is reliable enough now to measurably change the character of drunk-driving enforcement.

Charlottesville tested the sensor extensively last year in concert with researchers from the Insurance Institute. For six weeks, at Friday- and Saturday-night roadblocks, officers screened drivers who, if they were judged not to have been drinking by the police, then were approached by two medical technicians, wearing white laboratory coats to distinguish them from the police.

For three weeks, the test was conducted without the sensor and, for the other three weeks, the gadget was used. To induce cooperation, drivers who were found to be under the influence by the technicians but who had escaped detection by the police were not charged but were, instead, offered free transportation home. Almost all of the drivers cooperated. A total of 1,731 cars was checked both by police and the medical technicians.

While Rodenizer had high praise for the sensor, he noted that none of the Virginia cases that resulted in arrest has yet entered court appeals.

The analysis noted, however, that the sensor may only be used as an added part of established police procedures in which officers must have some reason to stop a car on the highway, to begin with, or at drunk-driving roadblocks, which have come into increasing use across the country in the last two years. The California Highway Patrol and several local law enforcement agencies experimented with roadblocks during the Christmas and New Year’s holiday season last year.

In California, however, there could be one additional legal barrier to use of the sensor, the Highway Patrol said--though both the patrol and the Los Angeles Police Department expressed interest in the sensor and both agencies said they would be willing to test it under street conditions.

The problem is that California has what lawyers call an “implied consent” provision in its drunk-driving law. Under the state vehicle code, any licensed driver must consent to giving a breath, blood or urine sample if he or she is legally asked to do so.

But the law requires the driver to furnish only one sample. Maurice Hannigan, a Highway Patrol deputy commissioner, said a formal legal ruling would probably be necessary here to be certain the sniffing of the sensor did not constitute the single official sample a driver may be compelled to provide.