Home
Please make a copy of this document and include this in your GitHub repository for your submission, using the tag #AndroidDevChallenge
I'm trying to solve a simple problem of finding a parking spot near me, currently focusing for SFO. There are many studies which show how much time an average person spends trying to find a parking spot. Finding a spot is like hunting for a needle in a haystack. When you do find one, it’s hard not to approach it with trepidation- Why is this spot empty?
Prototype: By looking at parking boards(Most of the time user is confused & unable to decide or not aware of the signs provided which could sometimes lead to ticket), determine if i can park here or not.
- Getting inputs from user until we have enough data, like image of board, asking questions, is it metered, public, private or luxury parking also like number of hours I can park here.
- User using application can mark spot parked or not parked (OR) by getting GPS input like Google's car parked last time & by checking is that user left that spot.
- Live on map available spots to park, planning to use Google cloud, once we have all the above accomplished we would have enough data to mark it on a map.
- With all information & data collected providing user best possible & nearest location to park without spending much time searching for one.
Currently for prototype I'm planning to use data from kaggle to train model with relevant parking board images, which could predict whether a user can park or not & with with android application using camera to predict parking spot & background with machine learning model.Describe in 250 words what the feature or service will do and how you’ll use Machine Learning to push the bar:
Project currently in exploring & learning with existing object detection project provided on tensorflow, modifying model as per project needs.
- Project planning to meet goals.
- Guidance with Machine learning & Model optimization & improvement.
- Guidance on integrating Google Maps, how to take most out of kaggle & tensorflow.
- Guidance with system design & building the project.
- Guidance on testing the project.
- November 2019: Exploring & learning with existing object detection project provided on tensorflow, modifying model as per project needs.
- December 2019: Begin the model training process & testing prediction of parking spots. Prototype would be ready by the end of the month.
- January 2020: Implementation of the Machine Learning to android application. Adding new features & services.
- February 2020: Optimizations & alpha testing with users, friends & family.
- March 2020: Fine-tune the app and re-train the models, in order to achieve the highest accuracy possible. Start beta testing phase.
- April 2020: Final bug fixes and polishing for the final release..
Describe where your project is, how you could use Google’s help in the endeavor, and how you plan on using On-Device ML technology to bring the concept to life. The best submissions have a great idea combined with a concrete path of where you plan on going, which should include: (1) any potential sample code you’ve already written, (2) a list of the ways you could use Google’s help, (3) as well as a timeline on how you plan on bringing it to life by May 1, 2020.
Currently I’m a Data Science Scholarship student at Udacity & also completed Android Nanodegree at udacity, love to add more skills to my bucket list & a lifelong student. Previously I was a triage engineer & also android application developer.Here is my linkedIn profile : https://www.linkedin.com/in/archanashivashankar/ A great idea is just one part of the equation; we also want to learn a bit more about you. Share with us some of your other projects so we can get an idea of how we can assist you with your project.