<p>In just seven months this year, Bengaluru Traffic Police have registered over 4.5 lakh cases against mandatory seat belt rule violations. This trumps the highest number of cases booked in the past six years — 3,08,145, in 2020. The rise in the number, experts said, could also indicate the successful adoption of AI technologies to track these violations.</p>.<p>A total of 4,54,972 cases were registered against motorists driving without a seat belt until July 31. This is a whopping 270.11% increase from the previous year, an alarming jump even before the year is to end.</p>.Bengaluru: Basic life support training for traffic police.<p>In January, DH had reported on the anomaly observed in 2022, which saw the lowest number of recorded seat belt violations since 2017, at 1,22,929 cases, even while the number of traffic violations kept increasing. The 2022, figures showed a 60% decrease from 2020, which reported the previous high in violations.</p>.<p>There could be two plausible reasons for the alarming increase in the number of cases this year — a growing number of callous road users or greater efficiency of the AI-powered cameras under the Intelligent Traffic Management System, installed at 50 crucial traffic junctions. Experts appear to favour the second possibility.</p>.Seat-belt violations at 5-yr low in 2022 in Bengaluru.<p>S N Omkar, chief research scientist at the Indian Institute of Science, said the numbers are rising because of the accuracy of AI models behind the camera systems at these junctions. “These models follow a very basic algorithm; so, the possibility for error is very low,” he said.</p>.<p>Prof Ashish Verma, convenor, Sustainable Transportation Lab at IISc, echoed the argument. Cameras capture more cases than earlier, with greater accuracy than a non-automated physical checking mechanism, he noted. “There are limited chances of error in these AI-powered cameras because they are trained to capture specific violations. These high numbers are actually a good example of why we need 24*7 enforcement using technology because they capture every violation without bias,” he said.</p>.<p>The cameras capture between 2,000 and 2,500 cases of seat belt violations every day. They have achieved an almost 99% accuracy rate, noted traffic police chief M N Anucheth. “There are some cases where violations are not booked if the colour of the belt is the same as the clothing, or when it is dark, but the margin of error is very low; maybe one in 10,000 cases is missed,” he said.</p>
<p>In just seven months this year, Bengaluru Traffic Police have registered over 4.5 lakh cases against mandatory seat belt rule violations. This trumps the highest number of cases booked in the past six years — 3,08,145, in 2020. The rise in the number, experts said, could also indicate the successful adoption of AI technologies to track these violations.</p>.<p>A total of 4,54,972 cases were registered against motorists driving without a seat belt until July 31. This is a whopping 270.11% increase from the previous year, an alarming jump even before the year is to end.</p>.Bengaluru: Basic life support training for traffic police.<p>In January, DH had reported on the anomaly observed in 2022, which saw the lowest number of recorded seat belt violations since 2017, at 1,22,929 cases, even while the number of traffic violations kept increasing. The 2022, figures showed a 60% decrease from 2020, which reported the previous high in violations.</p>.<p>There could be two plausible reasons for the alarming increase in the number of cases this year — a growing number of callous road users or greater efficiency of the AI-powered cameras under the Intelligent Traffic Management System, installed at 50 crucial traffic junctions. Experts appear to favour the second possibility.</p>.Seat-belt violations at 5-yr low in 2022 in Bengaluru.<p>S N Omkar, chief research scientist at the Indian Institute of Science, said the numbers are rising because of the accuracy of AI models behind the camera systems at these junctions. “These models follow a very basic algorithm; so, the possibility for error is very low,” he said.</p>.<p>Prof Ashish Verma, convenor, Sustainable Transportation Lab at IISc, echoed the argument. Cameras capture more cases than earlier, with greater accuracy than a non-automated physical checking mechanism, he noted. “There are limited chances of error in these AI-powered cameras because they are trained to capture specific violations. These high numbers are actually a good example of why we need 24*7 enforcement using technology because they capture every violation without bias,” he said.</p>.<p>The cameras capture between 2,000 and 2,500 cases of seat belt violations every day. They have achieved an almost 99% accuracy rate, noted traffic police chief M N Anucheth. “There are some cases where violations are not booked if the colour of the belt is the same as the clothing, or when it is dark, but the margin of error is very low; maybe one in 10,000 cases is missed,” he said.</p>