WINNING the VISTA APAC HACKATHON 2019

2019 has been a year full of surprises and great moments. The biggest highlight of the year was undoubtedly the winning moment at the VISTA APAC Hackathon. As a reward, we were awarded a sizeable amount in winnings followed by a 20-minute walkthrough of our solutions with one of the greatest minds of this generation Robert F Smith.

Heres a quick picture of our Team presenting our Smart City solution to Robert CEO of Vista. It was truly an honor presenting to Robert and he had some great suggestions on how we can improve our solution and take it to market.

Winning Solution: Smart Accident Response

The genesis of our idea was to showcase the ability to save lives by predicting and detecting accidents in Real-time using AI. Over 3000 deaths occur each day due to Road-Fatalities and our solution focuses on the fact that we can prevent such fatalities by having an early warning detection system.

This would enable the city to send out Smart First responders to the scene of the accident to treat potential victims on the spot. A futuristic example of this could be leveraging a network of drones to help respond to incidents across a city detected by our solution. A city like Bombay would only need a network of 80 drones to cover the entire region.

The Technology

Broadly speaking we implemented 2 Key outcomes in our solution. First, we would predict the probability of an accident to occur at any intersection. We achieved this by leveraging traffic sensor data. These sensors are mostly already available across most cities but the data is not being leveraged by most cities. We were able to correlate both traffic sensor data with historic crashes that occurred in the city using Artificial Intelligence. The same AI model is scored against the traffic data in real-time to predict the probability of an accident to occur.

For example, our solution would raise an alert saying there is 70% probability of a high impact incident happening at an intersection in downtown DC in the next 20 min. This early-warning prediction would allow responders to alert traffic patterns or deploy law enforcement to help prevent incidents.

Secondly, our solution could analyze video feeds and detect accidents that occur. In essence, we implemented Computer Vision. This would enable our solution to provide a comprehensive command center-like dashboard to identify critical accidents in real-time.

• 

• 

Awards & the Team:

Our team of 5 was awarded both the Peer Recognition award along with the Best business Idea award. This was truly a team effort and we had to implement most of this solution in under 24 hours.