Safe-T

Safety for Clean Transport (E-Scooters, E-Skateboards, E-Bikes, Etc...)

Comment

Safe-T

We wanted to create a electric vehicle sharing application that allows users to share last mile transportation with their friends including locking features with facial recognition, and process a SafeT Score that determines whether the user should ride to their destination.

First we prototyped a custom electric skateboard built using lithium ion batteries soldered in parallel and series along with a pcbboard with mofsets to act as a speed controller. We added motors to the skateboard and controlled the skateboard using a bluetooth remote. The board has regenerative braking as well to regenerate energy.

Next, we built a model on google automl and trained the machine to recognize faces. Then after hours of multiple training and labeling images, we created a rest endpoint so servers can create requests to check whether a face passes the unlocking test. We used python with opencv and pybluez to take real-time images using a camera to check whether the face is valid and then pybluez module to connect via bluetooth and unlock the skateboard.

Lastly, we built a mobile app to be cross-platform (both web and mobile) to determine road conditions using Google Maps API and LA Open Data for roads for iOS using swift and a multitude of modules such as alamofire for requests.

Goal

SafeT for Safe Transportation. We dream of a cleaner, safer tomorrow for all citizens within the LA area. Our project will support the expanding industry of clean electric transports, from a home built electric skateboard to one of those rental scooters, our platform will allow anyone to analyze the viability of using E-transport to get from one place to another.

Technology

  • Map analysis and information built on Google Maps API
  • Custom built electric skateboard for built in Bluetooth reception
  • Custom built electric skateboard with regenerative braking and speeds of up to 25 mph
  • Custom trained model built and trained on Google Cloud AutoML
  • Python predict generates a rating out of 1 based on it's prediction
  • Face recognition to allow users to unlock/lock their vehicles

Safe-T: Cross Platform

iOS app for scanning locations in the greater LA area for multiple criteria in determining the safety of electric transport in going from location to destination. Using the following criteria, our application will determine a SafeT score for your proposed ride across the given route.

  • Criminal Activity: Taken from the LA Crime Database, we find out how much crime there is along your path and judge the safety of that route, e-transport generally not as safe as car
  • Road Quality: Taken from LA Open Data, we find out the quality of roads along your path, as e-transport is generally more susceptible to bumpiness
  • Distance: Electric scooters and skateboards are limited by the range of their battery, combined with their slower speed, distance is a limiting factor

Combined with the web app, we have a full package product that both manages the safety of your vehicle with face unlocking, and the safety of your body with our not patented SafeT Score.

Share this project:

Updates