Alcohol represents one of the most recurrent risk factors in traffic accidents. According to the General Directorate of Traffic (DGT), alcohol consumption is related to between 30% and 50% of fatal accidents, with significant repercussions that translate into approximately 1,500 fatalities and another 50,000 people injured each year in Spanish roads.
The campaigns launched by the authorities, such as the memorable spot starring Stevie Wonder in 1994, under the slogan 'If you drink, don't drive', and the preventive controls carried out by traffic officers on the road have not managed to eradicate a practice. that continues to claim lives and generate tragic consequences daily.
Today, there are still too many drivers who get behind the wheel under the influence of alcohol without being aware of the risk that this represents both for their safety and that of other road users. Although it is best not to drink alcoholic beverages if you have to drive, the DGT establishes a blood alcohol limit for those who choose to do so.
To evaluate their blood alcohol level, drivers have the option of using a breathalyzer similar to those used by traffic officers to carry out preventive checks. Although it is a device within reach of all budgets - the cheapest breathalyzers are available on the market for less than 10 euros - there are few drivers who carry a gadget like this in the car.
In order to provide an additional and more accessible tool, researchers at Stanford University (California, United States) and Toronto (Canada) are moving towards integrating breathalyzer detection functions into mobile phones. It consists of implementing an application on smartphones based on an algorithm that measures the different tones of voice according to the alcohol ingested.
This revolutionary method measures the difficulty a person has in reading a tongue twister fluently and understandably. Once the reading is recorded, this audio is compared with a previous one recorded when that person had not ingested alcohol. By collating the results, the algorithm determines the blood alcohol level according to the degree of complexity for the reading.
The first tests carried out by the researchers were carried out on a group of 18 people, men and women aged 21 to 62, who were given a dose of alcohol according to their weight. Each of them were assigned various tongue twisters that they had to read before starting the experiment and then every hour for 7 hours after drinking.
The researchers examined different parameters related to reading ability, to determine the effect of alcohol on diction and, consequently, the impact on their reflexes and attention span. The results obtained were compared with the breath alcohol tests that were carried out simultaneously, offering an accuracy of 98% between both.
Although the experiment has proven to be reliable, the researchers recognize the need to carry out numerous tests to confirm the validity of voice patterns as an effective tool for detecting blood alcohol levels. They also admit that other variables, such as the way you walk or the way you compose a text message, could be used to obtain even more precise results.
