This past weekend I got the opportunity and had the pleasure of participating in my first hackathon at the UofTHacks VI, which was at the University of Toronto campus from January 18th to the 20th at the Bahen Centre for Information Technology. The event consisted of over 500 hackers, and spanned 36 hours, and would ultimately prize 3 winners.
An event, typically lasting several days, in which a large number of people meet to engage in collaborative computer programming. We look past this traditional definition of a hackathon and take it to new heights. UofTHacks is the University of Toronto’s annual hackathon. At UofTHacks, we bring people together to make new friends, learn new things, and bring new ideas to life.
I was participating as a Microsoft sponsor. This meant I was available to help mentor the students who had questions on Microsoft Azure or anything for that matter, and provide them with an insider perspective of how things work in the industry. I also then got to be a judge for the Microsoft prize which was based on projects that leverage Microsoft Azure services.
It was a lot of fun to work with these students as they had such drive and passion for what they worked on. The judging aspect was probably the best part as I got to take a look at over a dozen projects and see what they built, why they built it, how they built and then where would they take it next. Almost all the projects had real world applications and some even created business plans as part of their sales pitch.
Some of the big sponsors at the event were Microsoft, Google, IBM, Interact, Intel, Standard Library, ChainSafe, Lyft and SmartCar, plus many others.
As for the Microsoft prize, it was 4 Xbox One X 1TB consoles. This was presented to team Blink who created an application that would sit in your car and monitor the drivers eyes for signs of fatigue, drowsiness, or any other impairment. They leveraged the Microsoft Custom Vision Service and built their own data model that consisted over 250 images they gathered from the event of peoples eyes. They eyes were that of different ages, different skin colors, with and without glasses. They also retrained this model a number of times over the course of the 36 hour event, including the morning of judging session when the room had different lighting conditions. Their was a mobile and server components of the app that would send out SMS alerts to the registered emergency contact in the event impairment was detected and also alert emergency services. It was a well put together hack that had a clear application for issues we encounter today.
Here are some pictures of the event
This might of been my first hackathon but it definitely won’t be my last. I will keep an eye out for something more suited for the working developer that is maybe limited to a day, otherwise I hope to come back next year and help these students out and see what they create.
Team Blink – A machine learning eye detection app which determines if the driver of a smartcar is impaired & alerts 1st responders