Published: 25th November 2019
This Hyderabad start-up's software helps detect car speed, enables automatic number plate recognition. Here's how
The software is termed as a comprehensive traffic enforcement system and this technology can be used across Hyderabad and is several times cheaper than the conventional speed detection devices
It is a software developed by a Hyderabad start-up that helped nail the fact accurately that the accused involved in the Biodiversity flyover accident was driving at a speed of 105 kilometres per hour. This information has been crucial in the case as without that it was highly likely to assume it was only the 'S' shape of the flyover, which was the main culprit in the accident.
The start-up, who helped the police with this technology is AEye Softlabs, who are presently being incubated in T-Hub at IIIT Hyderabad. The Cyberabad Police tied up with this start-up earlier this year and have been hand-holding them to develop a tech that helps reduce the intervention of traffic police to detect traffic violations.
The 'a-eye' (artificial eye) has developed a software that automatically calculates the speed of the cars on the flyover using the CCTV footage from various points on the road. In case of the Biodiversity flyover, they calculated how much time it took for the car to travel from point A to B and calculated the speed at which it could be driving. The software does this in a matter of a few seconds for every vehicle passing by. "It was this software that has helped us to nail all those who speeding on Biodiversity flyover. It helped us generate 550+ challans in a matter of a week without any speed laser guns. This technology is even more useful because on the flyover there is no central median or no footpath for our officers to stand, so just with CCTV camera we can pin those who over speeding," said DCP Vijay Kumar, Cyberabad Traffic Police.
The creators of the software call it a comprehensive traffic enforcement system and have said that this technology can be used across Hyderabad and is several times cheaper than the conventional speed detection devices. "We are working in collaboration with the police for the past 5 to 6 months and our area of concentration has been traffic violation-related acts. Earlier the personnel sitting in the Commisionerate would monitor the video of any kind of violation, capture and then generate a challan. We have completely automated that process while using the cameras which were already being used by them. We just had to install our software into it. Whenever a violation takes place — wrong route detection, no helmet, triple riding — the camera will automatically capture an image and then send a message to the violator on their phones. It's already running at multiple places in Hyderabad," said Uttam Byragoni, Founder and CEO of AEye Softlabs.
The founder explains that the software has made the task easier for the police as it was difficult to put laser guns on the flyover as they are narrow and they don't have space for someone to sit and do the job. Also, adding that the laser guns usually have an error margin of +5/-5 kilometre while calculating the speed, whereas the start-up's software provides a more accurate reading. "We issued around 550 traffic violation challans only the last one week, reducing the total number of traffic violations dramatically. All the violators receive it on their phones as an SMS notification. After the installation of the software, the police do not have to manually sit and issue challans," explains Uttam.
AEye Softlabs first began working with private enterprises with automatic number plate recognition a year and a half ago. "Our software can recognise vehicle numbers, all types- bike, lorry, trucks, cars with high accuracy. We use Artificial Intelligence to do this, that's why we are able to read numbers pretty quickly in a matter of milliseconds and even on difficult number plates that might not be readable by any other software," adds Uttam.
At first, they were only working in the private sector, like in hospitals, hotels, IT parks, companies wherein there would be security guards who would write down the car numbers as and when the vehicles came in and left a building. "However, after installing our software it became fully automated and no one has to manually do it day and night," he says. He concludes by saying that the software is beneficial for Tier 2 and Tier 3 cities where traffic cameras are already running and they just have to install the software at no extra cost that they would have spent on high-end equipment.
(With inputs from Donita Jose, TNIE)