While brainstorming for MakeUofT, I happened to look outside. I noticed how icy the sidewalks and roads were and thought about the dangers it could pose on pedestrians and road users. Moreover, while effort had been made so salt some parts of the sidewalks, there was a lot if inconsistency in this with some parts being over-salted while other patches of ice remained unsalted. This was how Icety was created.
What it does
Icety is an automated vehicle salt-spreader that regulates the amount of salt distributed based on the temperature, and the amount of salt already on the road that it detects using computer vision. This detection is done using Microsoft Azure Vision.
How we built it
We used an Arduino Uno Rev3 microcontroller board connected to a Windows laptop and breadboard.
Challenges we ran into
We originally tried to program the Arduino from scratch to power the motor, but then learned that there was a pre-existing library within the IDE made for working with hobbyist motors. Also, we were unable to figure out how best to detect salt on the roads or differentiate between snowy and icy roads, but we then simplified the problem by focusing on the specific salt that is used on the University of Waterloo campus with a bluish-green hue.
Accomplishments that we're proud of
We were able program the arduino and create a twilio platform for pedestrian users.
What we learned
How to narrow down multiple possible design options and select the best one for out design. We also learn about how image identification works in Azure.
What's next for ICETY
Exploring more of the technologies we used to be able to create a real working model that can be used on vehicles in the city.