We were inspired by the Starship Robots themselves. They are an exciting new addition to campus, and we wanted to facilitate their growing popularity with students by our website, Cockbots. Cockbots allows users to connect and bond over the Starship Robots, and solely these robots.
Cockbots is a social media website where users can only post pictures of the Starship Robots. We have an AI algorithm that detects if a photo contains a bot or not, which denies any picture that is detected to not contain a bot. Cockbots additionally contains several features common to social media, such as users and profile pictures, as well as posts and likes. Users can post any amount of bot pictures they would like and can view a feed of previous bot pictures posted by other individuals as well.
Cockbots is ran using XAMPP, a free and open-source server solution stack package. It utilizes a MySQL database with three different tables, for users, posts, and likes. This was constructed using a variety of PHP, HTML, CSS, and JS. Additionally, Python and TensorFlow were used to make and implement the AI model for detecting bots.
Some of the challenges we ran into regarding Cockbots is proper Python execution in tandem with the website itself. There were issues file pathing to the pictures as well as the Python files themselves. The database was corrupted several times, leading to work loss and necessary backtracking and reconstruction. However, despite these challenges, we made a cohesive project working together as a team to venerate those little robots.
We learned a lot about the intricacies of social media websites, and how many features are necessary to make it functional for users. Additionally, how much of an issue it can be to make those features interact with and enhance each other. We learned, on the web side, how to create servers and manage users as well as their data across multiple databases. On the AI side, we learned the basics of training an AI detection model, as well as implementing it into a website through Python.
Some additional features we have planned for Cockbots in the future is a comment system so that users can interact with each other on the same post. Web security is also a priority, as we want our users to feel secure while browsing for robots. The AI model could be better at detecting robots, and thus that is another part of the product that can be improved upon. Lastly, one big part is that Cockbots would do excellent as a mobile app, and that is an avenue that we should take advantage of and develop one ourselves.
Log in or sign up for Devpost to join the conversation.