Published: 23rd May 2017
This sleep safe detection system by two engineering students may actually bring down the number of road accidents. Here's how
Ferdinand Thomas and Jacob Sabu who are second- year students at the Model Engineering College, Thrikkakkara in Kochi, Kerala developed a system that can alert sleeping drivers
As per the statistics put out by Kerala Police, the number of deaths caused by the mistakes of the driver stands at a staggering 38,189 in the year 2016. In most cases, it's the drowsiness of the driver that plays the villain. A system developed by two students - Ferdinand Thomas and Jacob Sabu - of Model Engineering College, Thrikkakkara, Kochi, takes this aspect into consideration - it could identify the drowsiness in a driver and alert him thereby avoiding an accident.
This system with basic hardware requirements is way cheaper than the automatic systems in cars of Tesla or Volvo. However, it can help you reach your destination safe and sound
The design was a big hit at the IBeTo Hackathon that was held in the last two days (a 24-hour continuous event) of April at Kerala Startup Mission in Kochi and won the first prize.
"Many people we knew have lost their lives to accidents. We learned that most of it were caused because the driver drowned off for a second or two. So when we signed in for the hackathon and was given the topic 'Revolutionizing Road Safety', we decided to make a system that could offset the chances of accidents caused due to drowsiness," says Thomas and Sabu, second-year students of Electronics and Communication Engineering.
Saving Lives: Student duo at Model Engineering College invent the Sleep Safe Detection System
The duo named their system Sleep Safe Detection System. It uses image processing software to detect drowsiness in a driver.
They explain their system as follows:
1.A camera is mounted near the steering wheel and the driver is video graphed throughout the journey.
2.The video feed is analysed in real time using raspberry pi with Image processing to detect facial changes of the driver like blinking of eyes, yawning etc
3.If the symptoms are detected for more than 4 seconds, the processor is notified
4.The processor alerts the driver of his state of attentiveness by preprogrammed voice notification.
5. The driver loses manual control of the car and automatic parking is initiated by the system. (If the triggering of automated parking was initiated incorrectly, a switch is provided at the drivers' end to reset the system instantly and manual control is regained)
6.Automatic parking utilises ultrasonics sensors we usually use as reverse parking sensors and cameras for higher precision. Cameras are used to calculate current traffic to avoid accidents due to sudden parking
7.After parking, the driver can rest for a while or reset the system and continue the trip.
The duo says the system is currently programmed for perfect road conditions, which is non-existent. "If the system has to be put to use on real world roads, it requires better and refined intelligence. In our next phase we intend to learn about the types of roads and how to adapt auto parking to all road conditions," they say. This system with basic hardware requirements is way cheaper than the automatic systems in cars of Tesla or Volvo. However, it can help you reach your destination safe and sound.
The IBeTo Hackathon was held to develop a working prototype based on the problem statements. These prototypes can further be iterated and be launched into the market. In addition to the prize money, the top 3 contestants have got direct entry into the IBeTo finals which guarantees them proper mentorship and guidance from the experts in their field.
Improving Tomorrow: Two students from Model Engineering college have been listed as finalists at IBeto 2017
Thomas and Sabu have been selected as finalists for the IBeto 2017 to be held in September. They say they intend to upgrade the product in the meantime and present it at the venue with the new additions.